Hit enter to search or ESC to close

Paper Digest: ICASSP 2020 Highlights

admin

Download ICASSP-2020-Paper-Digests.pdf– highlights of all ICASSP-2020 papers.

The IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) is one of the top signal processing conferences in the world. In 2020, it is to be held virtually between May 4 and May 8, 2020.

To help the community quickly catch up on the work presented in this conference, Paper Digest Team processed all accepted papers, and generated one highlight sentence (typically the main topic) for each paper. Readers are encouraged to read these machine generated highlights / summaries to quickly get the main idea of each paper.

If you do not want to miss any interesting academic paper, you are welcome to sign up our free daily paper digest service to get updates on new papers published in your area every day. You are also welcome to follow us on Twitter and Linkedin to get updated with new conference digests.

Paper Digest Team
team@paperdigest.org

TABLE 1: ICASSP 2020 Papers

Title Authors Highlight
1 Addressing The Confounds Of Accompaniments In Singer Identification T. Hsieh, K. Cheng, Z. Fan, Y. Yang and Y. Yang In this paper, we attempt to address this issue. Specifically, we employ open-unmix, an open source tool with state-of-the-art performance in source separation, to separate the vocal and instrumental tracks of music. We then investigate two means to train a singer identification model: by learning from the separated vocal only, or from an augmented set of data where we “shuffle-and-remix” the separated vocal tracks and instrumental tracks of different songs to artificially make the singers sing in different contexts.
2 Disentangled Multidimensional Metric Learning for Music Similarity J. Lee, N. J. Bryan, J. Salomon, Z. Jin and J. Nam To do so, we adapt a variant of deep metric learning called conditional similarity networks to the audio domain and extend it using track-based information to control the specificity of our model.
3 Learning the Helix Topology of Musical Pitch V. Lostanlen, S. Sridhar, B. McFee, A. Farnsworth and J. P. Bello This article addresses the problem of discovering this helical structure from unlabeled audio data.
4 Audio-Based Auto-Tagging With Contextual Tags for Music K. M. Ibrahim, J. Royo-Letelier, E. V. Epure, G. Peeters and G. Richard In this work, we study the relationship between user context and audio content in order to enable context-aware music recommendation agnostic to user data. Using this, we create and release a dataset of ~50k tracks labelled with 15 different contexts.
5 Accurate and Scalable Version Identification Using Musically-Motivated Embeddings F. Yesiler, J. Serr? and E. G?mez In this paper, we present MOVE, a musically-motivated method for accurate and scalable version identification.
6 Similarity Learning For Cover Song Identification Using Cross-Similarity Matrices of Multi-Level Deep Sequences C. Jiang, D. Yang and X. Chen In this paper, we propose a new Siamese network architecture for music melody similarity metric learning.
7 Two-Step Sound Source Separation: Training On Learned Latent Targets E. Tzinis, S. Venkataramani, Z. Wang, C. Subakan and P. Smaragdis In this paper, we propose a two-step training procedure for source separation via a deep neural network.
8 A Multi-Phase Gammatone Filterbank for Speech Separation Via Tasnet D. Ditter and T. Gerkmann In this work, we investigate if the learned encoder of the end-to-end convolutional time domain audio separation network (Conv-TasNet) is the key to its recent success, or if the encoder can just as well be replaced by a deterministic hand-crafted filterbank.
9 Improving Voice Separation by Incorporating End-To-End Speech Recognition N. Takahashi, M. K. Singh, S. Basak, P. Sudarsanam, S. Ganapathy and Y. Mitsufuji In this work, we propose to explicitly incorporate the phonetic and linguistic nature of speech by taking a transfer learning approach using an end-to-end automatic speech recognition (E2EASR) system.
10 Dual-Path RNN: Efficient Long Sequence Modeling for Time-Domain Single-Channel Speech Separation Y. Luo, Z. Chen and T. Yoshioka In this paper, we propose dual-path recurrent neural network (DPRNN), a simple yet effective method for organizing RNN layers in a deep structure to model extremely long sequences.
11 Controlling the Perceived Sound Quality for Dialogue Enhancement With Deep Learning C. Uhle, M. Torcoliv and J. Paulus We propose a method for controlling the trade-off between the attenuation of the interfering background signal and the loss of sound quality.
12 Unsupervised Training for Deep Speech Source Separation with Kullback-Leibler Divergence Based Probabilistic Loss Function M. Togami, Y. Masuyama, T. Komatsu and Y. Nakagome In this paper, we propose a multi-channel speech source separation method with a deep neural network (DNN) which is trained under the condition that no clean signal is available.
13 A Framework for the Robust Evaluation of Sound Event Detection ?. Bilen, G. Ferroni, F. Tuveri, J. Azcarreta and S. Krstulovic This work defines a new framework for performance evaluation of polyphonic sound event detection (SED) systems, which overcomes the limitations of the conventional collar-based event decisions, event F-scores and event error rates.
14 Weakly-Supervised Sound Event Detection with Self-Attention K. Miyazaki, T. Komatsu, T. Hayashi, S. Watanabe, T. Toda and K. Takeda In this paper, we propose a novel sound event detection (SED) method that incorporates a self-attention mechanism of the Transformer for a weakly-supervised learning scenario.
15 A Sequence Matching Network for Polyphonic Sound Event Localization and Detection T. N. Tho Nguyen, D. L. Jones and W. Gan We propose a two-step approach that decouples the learning of the sound event detection and directional-of-arrival estimation systems.
16 Few-Shot Acoustic Event Detection Via Meta Learning B. Shi, M. Sun, K. C. Puvvada, C. Kao, S. Matsoukas and C. Wang We study few-shot acoustic event detection (AED) in this paper.
17 Few-Shot Sound Event Detection Y. Wang, J. Salomon, N. J. Bryan and J. Pablo Bello In this work, we (1) adapt state-of-the-art metric-based few-shot learning methods to automate the detection of similar-sounding events, requiring only one or few examples of the target event, (2) develop a method to automatically construct a partial set of labeled examples (negative samples) to reduce user labeling effort, and (3) develop an inference-time data augmentation method to increase detection accuracy.
18 Sound Event Detection in Synthetic Domestic Environments R. Serizel, N. Turpault, A. Shah and J. Salamon We present a comparative analysis of the performance of state-of-the-art sound event detection systems.
19 Learning to Separate Sounds from Weakly Labeled Scenes F. Pishdadian, G. Wichern and J. L. Roux We propose objective functions and network architectures that enable training a source separation system with weak labels.
20 Improving Universal Sound Separation Using Sound Classification E. Tzinis, S. Wisdom, J. R. Hershey, A. Jansen and D. P. W. Ellis In this paper, we utilize the semantic information learned by sound classifier networks trained on a vast amount of diverse sounds to improve universal sound separation.
21 Source Separation with Weakly Labelled Data: an Approach to Computational Auditory Scene Analysis Q. Kong, Y. Wang, X. Song, Y. Cao, W. Wang and M. D. Plumbley In this work, we propose a source separation framework trained with weakly labelled data.
22 Boosted Locality Sensitive Hashing: Discriminative Binary Codes for Source Separation S. Kim, H. Yang and M. Kim In this study, we propose an adaptive boosting approach to learning locality sensitive hash codes, which represent audio spectra efficiently.
23 A Frequency-Domain BSS Method Based on l1 Norm, Unitary Constraint, and Cayley Transform S. Emura, H. Sawada, S. Araki and N. Harada We propose a frequency-domain blind source separation method that uses (a) the l1 norm of orthonormal vectors of estimated source signals as a sparsity measure and (b) Cayley transform for optimizing the objective function under the unitary constraint in the Riemannian geometry approach.
24 End-To-End Non-Negative Autoencoders for Sound Source Separation S. Venkataramani, E. Tzinis and P. Smaragdis In this paper, we generalize NMF to develop end-to-end non-negative auto-encoders and demonstrate how they can be used for source separation.
25 Coincidence, Categorization, and Consolidation: Learning to Recognize Sounds with Minimal Supervision A. Jansen et al. With this motivation, we present a learning framework for sound representation and recognition that combines (i) a self-supervised objective based on a general notion of unimodal and cross-modal coincidence, (ii) a clustering objective that reflects our need to impose categorical structure on our experiences, and (iii) a cluster-based active learning procedure that solicits targeted weak supervision to consolidate categories into relevant semantic classes.
26 Acoustic Scene Classification for Mismatched Recording Devices Using Heated-Up Softmax and Spectrum Correction T. Nguyen, F. Pernkopf and M. Kosmider In this paper, we introduce two calibration methods to tackle these challenges.
27 Limitations of Weak Labels for Embedding and Tagging N. Turpault, R. Serizel and E. Vincent In this paper, we formulate a supervised learning problem which involves weak labels. We create a dataset that focuses on the difference between strong and weak labels as opposed to other challenges.
28 Mt-Gcn For Multi-Label Audio Tagging With Noisy Labels H. Shrivaslava, Y. Yin, R. R. Shah and R. Zimmermann We propose two ontology-based graph construction methods, and conduct extensive experiments on the FSDKaggle2019 dataset.
29 Acoustic Scene Classification Using Deep Residual Networks with Late Fusion of Separated High and Low Frequency Paths M. D. McDonnell and W. Gao We investigate the problem of acoustic scene classification, using a deep residual network applied to log-mel spectrograms complemented by log-mel deltas and delta-deltas.
30 End-To-End Auditory Object Recognition Via Inception Nucleus M. Ebrahimpour, T. Shea, A. Danielescu, D. Noelle and C. Kello In this paper, we propose a novel end-to-end deep neural network to map the raw waveform inputs to sound class labels.
31 Translation of a Higher Order Ambisonics Sound Scene Based on Parametric Decomposition M. Kentgens, A. Behler and P. Jax This paper presents a novel 3DoF+ system that allows to navigate, i.e., change position, in scene-based spatial audio content beyond the sweet spot of a Higher Order Ambisonics recording.
32 Blaster: An Off-Grid Method for Blind and Regularized Acoustic Echoes Retrieval D. D. Carlo, C. Elvira, A. Deleforge, N. Bertin and R. Gribonval This work proposes a novel approach to blindly retrieve the off-grid timing of early acoustic echoes from a stereophonic recording of an unknown sound source such as speech.
33 Evaluation of Sensor Self-Noise In Binaural Rendering of Spherical Microphone Array Signals H. Helmholz, J. Ahrens, D. L. Alon, S. V. Amengual Gar? and R. Mehra We use the Real-Time Spherical Array Renderer (ReTiSAR) to analyze and auralize the propagation of sensor self-noise through the processing pipeline.
34 Mutual-Information-Based Sensor Placement for Spatial Sound Field Recording K. Ariga, T. Nishida, S. Koyama, N. Ueno and H. Saruwatari A sensor (microphone) placement method based on mutual information for spatial sound field recording is proposed.
35 Fast Acoustic Scattering Using Convolutional Neural Networks Z. Fan, V. Vineet, H. Gamper and N. Raghuvanshi We propose training a convolutional neural network to map from a convex scatterer?s crosssection to a 2D slice of the resulting spatial loudness distribution.
36 Frequency-Dependent Directional Feedback Delay Network B. Alary and A. Politis In this paper, we present a modified formulation of the Directional Feedback Delay Network method that allows both frequency-and direction-dependent reverberation.
37 Speech Enhancement Using Self-Adaptation and Multi-Head Self-Attention Y. Koizumi, K. Yaiabe, M. Delcroix, Y. Maxuxama and D. Takeuchi This paper investigates a self-adaptation method for speech enhancement using auxiliary speaker-aware features; we extract a speaker representation used for adaptation directly from the test utterance.
38 PEVD-Based Speech Enhancement in Reverberant Environments V. W. Neo, C. Evers and P. A. Naylor In this work, we focus on reverberant environments. It is shown that, by exploiting the lack of correlation between speech and the late reflections, further noise reduction can be achieved.
39 DNN-Based Speech Presence Probability Estimation for Multi-Frame Single-Microphone Speech Enhancement M. Tammen, D. Fischer, B. T. Meyer and S. Doclo In this paper, we propose to use a bi-directional long short-term memory deep neural network (DNN) to estimate the SPP for each TF bin.
40 Nonlinear Spatial Filtering for Multichannel Speech Enhancement in Inhomogeneous Noise Fields K. Tesch and T. Gerkmann In this paper, we show that a joint spatial-spectral nonlinear filter is not only advantageous for noise distributions that are significantly more heavy-tailed than a Gaussian but also for distributions that model inhomogeneous noise fields while having rather low kurtosis.
41 Generalized Coherence-Based Signal Enhancement H. W. L?llmann, A. Brendel and W. Kellermann This contribution presents a novel approach for coherence-based signal enhancement.
42 Speaker Independence of Neural Vocoders and Their Effect on Parametric Resynthesis Speech Enhancement S. Maiti and M. I. Mandel Here we propose to use the high quality speech generation capability of neural vocoders for better quality speech enhancement.
43 Robust and steerable kronecker product differential beamforming With rectangular microphone arrays G. Huang, J. Benesty, J. Chen and I. Cohen In this paper, we consider rectangular shapes of planar microphone arrays.
44 Jointly Optimal Dereverberation and Beamforming C. Boeddeker, T. Nakatani, K. Kinoshita and R. Haeb-Umbach To this end, this paper presents a new derivation of the convolutional beamformer that allows us to factorize it into a WPE dereverberation filter, and a special type of a (non-convolutional) beamformer, referred to as a weighted MPDR (wM-PDR) beamformer, without loss of optimality.
45 Exploiting Rays in Blind Localization of Distributed Sensor Arrays S. Wozniak and K. Kowalczyk In this paper, we focus on estimators for inferring the relative geometry of distributed arrays and sources, i.e. the setup geometry up to a scaling factor.
46 A Novel Method for Obtaining Diffuse Field Measurements for Microphone Calibration N. Akbar, G. Dickins, M. R. P. Thomas, P. Samarasinghe and T. Abhayapala We propose a straightforward and cost-effective method to perform diffuse soundfield measurements for calibrating the magnitude response of a microphone array.
47 Multi-Channel Speech Source Separation and Dereverberation With Sequential Integration of Determined and Underdetermined Models M. Togami In this paper, we propose a joint multi-channel speech source separation and dereverberation method in which multiple speech sources and late reverberation are separated in an unsupervised manner.
48 Fast and Stable Blind Source Separation with Rank-1 Updates R. Scheibler and N. Ono We propose a new algorithm for the blind source separation of acoustic sources.
49 Modeling Plate and Spring Reverberation Using A Dsp-Informed Deep Neural Network M. A. Mart?nez Ram?rez, E. Benetos and J. D. Reiss Based on digital reverberators that use sparse FIR filters, we propose a signal processing-informed deep learning architecture for the modeling of artificial reverberators.
50 Deep Autotuner: A Pitch Correcting Network for Singing Performances S. Wager, G. Tzanetakis, C. Wang and M. Kim We introduce a data-driven approach to automatic pitch correction of solo singing performances.
51 Perceptual loss function for neural modeling of audio systems A. Wright and V. V?lim?ki This work investigates alternate pre-emphasis filters used as part of the loss function during neural network training for nonlinear audio processing.
52 One-Shot Parametric Audio Production Style Transfer with Application to Frequency Equalization S. I. Mimilakis, N. J. Bryan and P. Smaragdis In this paper], [we present a method that facilitates this process by inferring appropriate audio effect parameters in order to make an input recording sound similar to an unrelated reference recording.
53 Speech-To-Singing Conversion in an Encoder-Decoder Framework J. Parekh, P. Rao and Y. Yang In this paper our goal is to convert a set of spoken lines into sung ones.
54 Tensorflow Audio Models in Essentia P. Alonso-Jim?nez, D. Bogdanov, J. Pons and X. Serra In this work, we present a set of algorithms that employ TensorFlow in Essentia, allow predictions with pre-trained deep learning models, and are designed to offer flexibility of use, easy extensibility, and real-time inference.
55 Anomalous Sound Detection Based on Interpolation Deep Neural Network K. Suefusa, T. Nishida, H. Purohit, R. Tanabe, T. Endo and Y. Kawaguchi To solve the issue, we propose an approach to anomalous detection in which the model utilizes multiple frames of a spectrogram whose center frame is removed as an input, and it predicts an interpolation of the removed frame as an output.
56 A-CRNN: A Domain Adaptation Model for Sound Event Detection W. Wei, H. Zhu, E. Benetos and Y. Wang This paper presents a domain adaptation model for sound event detection. We have collected and annotated a dataset in Singapore with two types of recording devices to complement existing datasets in the research community, especially with respect to domain adaptation.
57 SPIDERnet: Attention Network For One-Shot Anomaly Detection In Sounds Y. Koizumi, M. Yasuda, S. Murata, S. Saito, H. Uematsu and N. Harada We propose a similarity function for one-shot anomaly detection in sounds (ADS) called SPecific anomaly IDentifiER network (SPI- DERnet).
58 Sound Event Detection Via Dilated Convolutional Recurrent Neural Networks Y. Li, M. Liu, K. Drossos and T. Virtanen In this paper, we propose to use a dilated CRNN, namely a CRNN with a dilated convolutional kernel, as the classifier for the task of SED.
59 A Deep Neural Network-Driven Feature Learning Method for Polyphonic Acoustic Event Detection from Real-Life Recordings M. Mulimani, A. B. Kademani and S. G. Koolagudi In this paper, a Deep Neural Network (DNN)-driven feature learning method for polyphonic Acoustic Event Detection (AED) is proposed.
60 Weakly Labelled Audio Tagging Via Convolutional Networks with Spatial and Channel-Wise Attention S. Hong, Y. Zou, W. Wang and M. Cao In this paper, we propose a novel attention mechanism, namely, spatial and channel-wise attention (SCA).
61 A Study on the Transferability of Adversarial Attacks in Sound Event Classification V. Subramanian, A. Pankajakshan, E. Benetos, N. Xu, S. McDonald and M. Sandler Our work focuses on studying the transferability of adversarial attacks in sound event classification.
62 Propeller Noise Detection with Deep Learning M. Thomas, F. Lionel and D. Laurent In this paper, we propose a new model of underwater propeller noise as well as its optimal statistical detector for detecting the presence of propeller in acoustic signal.
63 Duration Robust Weakly Supervised Sound Event Detection H. Dinkel and K. Yu This work aims to investigate the performance impact of fixed-sized window median filter post-processing and advocate the use of double thresholding as a more robust and predictable post-processing method.
64 A Comparison of Pooling Methods on LSTM Models for Rare Acoustic Event Classification C. Kao, M. Sun, W. Wang and C. Wang This paper focuses on investigating the dynamics of LSTM model on AEC tasks.
65 An Ontology-Aware Framework for Audio Event Classification Y. Sun and S. Ghaffarzadegan To capture such dependencies between the labels, we propose an ontology-aware neural network containing two components: feed-forward ontology layers and graph convolutional networks (GCN).
66 Task-Aware Mean Teacher Method for Large Scale Weakly Labeled Semi-Supervised Sound Event Detection J. Yan, Y. Song, L. Dai and I. McLoughlin In this paper, we propose a task-aware mean teacher method using a convolutional recurrent neural network (CRNN) with multi-branch structure to solve the SED and AT tasks differently.
67 Wawenets: A No-Reference Convolutional Waveform-Based Approach to Estimating Narrowband and Wideband Speech Quality A. A. Catellier and S. D. Voran Building on prior work we have developed a no-reference (NR) waveform-based convolutional neural network (CNN) architecture that can accurately estimate speech quality or intelligibility of narrowband and wideband speech segments.
68 A Neural Network for Monaural Intrusive Speech Intelligibility Prediction M. B. Pedersen, A. Heidemann Andersen, S. H. Jensen and J. Jensen In the present work, we propose a neural network for monaural intrusive SIP.
69 Source Coding of Audio Signals with a Generative Model R. Fejgin, J. Klejsa, L. Villemoes and C. Zhou We consider source coding of audio signals with the help of a generative model.
70 Full-Reference Speech Quality Estimation with Attentional Siamese Neural Networks G. Mittag and S. M?ller In this paper, we present a full-reference speech quality prediction model with a deep learning approach.
71 Enhanced Method of Audio Coding Using CNN-Based Spectral Recovery with Adaptive Structure S. Shin, S. K. Beack, W. Lim and H. Park This study proposes an enhanced method of audio coding based on spectral recovery with an adaptive structure that yields improved sound quality compared with the previous method.
72 Audio Codec Enhancement with Generative Adversarial Networks A. Biswas and D. Jia Specifically, with these two classes of signals as examples, we demonstrate a technique for restoring audio from coding noise based on generative adversarial networks (GAN).
73 Efficient and Scalable Neural Residual Waveform Coding with Collaborative Quantization K. Zhen, M. S. Lee, J. Sung, S. Beack and M. Kim We propose a collaborative quantization (CQ) scheme to jointly learn the codebook of LPC coefficients and the corresponding residuals.
74 A Dual-Staged Context Aggregation Method towards Efficient End-to-End Speech Enhancement K. Zhen, M. S. Lee and M. Kim In this paper, we propose a densely connected convolutional and recurrent network (DCCRN), a hybrid architecture, to enable dual-staged temporal context aggregation.
75 A Recurrent Variational Autoencoder for Speech Enhancement S. Leglaive, X. Alameda-Pineda, L. Girin and R. Horaud This paper presents a generative approach to speech enhancement based on a recurrent variational autoencoder (RVAE).
76 Speakerfilter: Deep Learning-Based Target Speaker Extraction Using Anchor Speech S. He, H. Li and X. Zhang To effectively utilize anchor speech, we propose a multi-level feature extraction and seamlessly integrate the features into a speech separation model.
77 Tackling Real Noisy Reverberant Meetings with All-Neural Source Separation, Counting, and Diarization System K. Kinoshita, M. Delcroix, S. Araki and T. Nakatani In this paper, we first consider practical issues required for improving the robustness of the all-neural approach, and then experimentally show that, even in real meeting scenarios, the all-neural approach can perform effective speech enhancement, and simultaneously outperform state-of-the-art systems.
78 Time-Domain Audio Source Separation Based on Wave-U-Net Combined with Discrete Wavelet Transform T. Nakamura and H. Saruwatari We propose a time-domain audio source separation method using down-sampling (DS) and up-sampling (US) layers based on a discrete wavelet transform (DWT).
79 Auditory Model Based Subsetting of Head-Related Transfer Function Datasets S. Spagnol In this article a novel HRTF subset selection algorithm based on auditory-model vertical localization predictions and a greedy heuristic is outlined, designed to identify a representative HRTF subset from a catalogue including the three biggest public datasets currently available (373 HRTFs overall).
80 Impulse Response Data Augmentation and Deep Neural Networks for Blind Room Acoustic Parameter Estimation N. J. Bryan To address this, we propose an AIR augmentation method that can parametrically control the T60 and DRR, allowing us to expand a small dataset of real AIRs into a balanced dataset orders of magnitude larger.
81 Individual Distance-Dependent HRTFS Modeling Through A Few Anthropometric Measurements M. Zhang, X. Wu and T. Qu In this paper, a method for predicting individual distance-dependent HRTFs using a few anthropometric parameters is proposed.
82 Evaluation of Deep-Learning-Based Voice Activity Detectors and Room Impulse Response Models in Reverberant Environments A. Ivry, I. Cohen and B. Berdugo We consider five different models to generate RIRs, and five different VADs that are trained with the augmented training set.
83 A Minimal Personalization of Dynamic Binaural Synthesis with Mixed Structural Modeling and Scattering Delay Networks M. Geronazzo, J. Y. Tissieres and S. Serafin This paper provides a small set of essential parameters for a personalized and effective real-time auralization with headphones.
84 Sound Texture Synthesis Using RI Spectrograms H. Caracalla and A. Roebel This article introduces a new parametric synthesis method for sound textures based on existing works in visual and sound texture synthesis.
85 Time Domain Velocity Vector for Retracing the Multipath Propagation J. Daniel and S. Kitic We propose a conceptually and computationally simple form of sound velocity that offers a readable view of the interference between direct and indirect sound waves.
86 Acoustic Matching By Embedding Impulse Responses J. Su, Z. Jin and A. Finkelstein This paper introduces a deep learning solution for two parts of the acoustic matching problem.
87 Joint Estimation Of Acoustic Parameters From Single-Microphone Speech Observations D. Looney and N. D. Gaubitch To address the inherent interplay that exists between these parameters, which can hinder existing methods designed to estimate only a single parameter, we propose a data-driven solution to jointly estimate all three parameters using a convolutional neural network.
88 A Fast Reduced-Rank Sound Zone Control Algorithm Using The Conjugate Gradient Method L. Shi, T. Lee, L. Zhang, J. K. Nielsen and M. Gr?sb?ll Christensen In this paper, we propose a fast reduced-rank sound zone control algorithm using the conjugate gradient (CG) method.
89 An Empirical Study on Acoustic Feedback Path Across Hearing Aid Users M. Guo In this empirical study, we measured feedback paths on different users wearing a RITE style hearing aid fitted with different domes and micro molds.
90 Low Complexity NLMS for Multiple Loudspeaker Acoustic ECHO Canceller Using Relative Loudspeaker Transfer Functions O. Schwartz, E. A. P. Habets and S. Gannot In this paper, we present an normalized least mean square (NLMS) algorithm for a multi-loudspeaker case using relative loudspeaker transfer functions (RLTFs).
91 A Multichannel Kalman-Based Wiener Filter Approach for Speaker Interference Reduction in Meetings P. Meyer, S. Elshamy and T. Fingscheidt We extend an existing approach by integrating methods from acoustic echo cancellation to improve the estimation of the interferer (noise) components of the filter, which leads to an improved signal-to-interferer ratio by up to 2.1 dB absolute at constant speech component quality.
92 Primary Path Estimator Based on Individual Secondary Path for ANC Headphones J. Fabry and P. Jax In this contribution, we propose an estimator for the individual primary path based on a measurement of the individual secondary path.
93 Efficient Multichannel Nonlinear Acoustic Echo Cancellation Based on a Cooperative Strategy M. M. Halimeh and W. Kellermann While a common approach to address nonlinear distortions, emitted by multiple loudspeakers and observed by multiple microphones, is to use post-filtering techniques, this paper proposes a cooperative strategy to rather model and then cancel such distortions.
94 Active Control of Line Spectral Noise with Simultaneous Secondary Path Modeling Without Auxiliary Noise M. Hu and J. Lu In this paper, we further extend the analysis and theoretically prove that even for line spectral noise, it is feasible to realize online secondary path modeling using only the output of the control source while exerting active control simultaneously.
95 Robust Frequency-Domain Recursive Least M-Estimate Adaptive Filter For Acoustic System Identification H. He, J. Chen, J. Benesty and Y. Yu To identify acoustic systems in non-Gaussian and Gaussian noises, a robust frequency-domain recursive least M-estimate (FRLM) adaptive filtering algorithm is proposed
96 Nearest Kronecker Product Decomposition Based Normalized Least Mean Square Algorithm S. S. Bhattacharjee and N. V. George In this paper, we derive the Least Mean Square (LMS) versions of adaptive algorithms which take advantage of NKP decomposition, namely NKP-LMS and NKP Normalized LMS (NKP-NLMS) algorithms.
97 Joint Beamforming and Reverberation Cancellation Using a Constrained Kalman Filter With Multichannel Linear Prediction S. Hashemgeloogerdi and S. Braun We propose a constrained Kalman filter based multichannel linear prediction method to jointly perform denoising and dereverberation efficiently using an online processing algorithm.
98 Multi-Microphone Complex Spectral Mapping for Speech Dereverberation Z. Wang and D. Wang This study proposes a multi-microphone complex spectral mapping approach for speech dereverberation on a fixed array geometry.
99 Predicting Word Error Rate for Reverberant Speech H. Gamper, D. Emmanouilidou, S. Braun and I. J. Tashev In this paper we propose predicting ASR performance in terms of the word error rate (WER) directly from acoustic parameters via a polynomial, sigmoidal, or neural network fit, as well as blindly from reverberant speech samples using a convolutional neural network (CNN).
100 Automatic Lyrics Alignment and Transcription in Polyphonic Music: Does Background Music Help? C. Gupta, E. Yilmaz and H. Li In this work, we propose to learn music genre-specific characteristics to train polyphonic acoustic models.
101 Local Key Estimation In Classical Music Recordings: A Cross-Version Study on Schubert?s Winterreise H. Schreiber, C. Wei? and M. M?ller In this work, we approach local key estimation on a unique cross-version dataset comprising nine performances (versions) of Schubert?s song cycle Winterreise?a challenging scenario of high musical ambiguity and subjectivity.
102 Improving Music Transcription by Pre-Stacking A U-Net F. Pedersoli, G. Tzanetakis and K. M. Yi We propose to pre-stack a U-Net as a way of improving the polyphonic music transcription performance of various baseline Convolutional Neural Networks (CNNS).
103 Learning to Rank Music Tracks Using Triplet Loss L. Pr?tet, G. Richard and G. Peeters In this work, we propose a method for direct recommendation based on the audio content without explicitly tagging the music tracks.
104 Transformer VAE: A Hierarchical Model for Structure-Aware and Interpretable Music Representation Learning J. Jiang, G. G. Xia, D. B. Carlton, C. N. Anderson and R. H. Miyakawa To achieve these two goals simultaneously, we designed the Transformer Variational AutoEncoder, a hierarchical model that unifies the efforts of two recent breakthroughs in deep music generation: 1) the Music Transformer and 2) Deep Music Analogy.
105 A Comparative Study of Western and Chinese Classical Music Based on Soundscape Models J. Fan, Y. Yang, K. Dong and P. Pasquier In this study, we examine whether we can analyze and compare Western and Chinese classical music based on soundscape models.
106 Audio-Based Detection of Explicit Content in Music A. Vaglio, R. Hennequin, M. Moussallam, G. Richard and F. d?Alch?-Buc We present a novel automatic system for performing explicit content detection directly on the audio signal.
107 New Metrics for Evaluating the Accuracy of Fundamental Frequency Estimation Approaches in Musical Signals J. Devaney This paper evaluates the magnitude of accuracy differences between a simple mean-based frame-level accuracy measurement and four metrics that capture more perceptually-relevant aspects of the evolution of f0 traces over time (perceived pitch, vibrato rate, vibrato depth, and jitter) for two score-informed f0 estimation algorithms.
108 Data-Driven Harmonic Filters for Audio Representation Learning M. Won, S. Chun, O. Nieto and X. Serrc We introduce a trainable front-end module for audio representation learning that exploits the inherent harmonic structure of audio signals.
109 Learning a Representation for Cover Song Identification Using Convolutional Neural Network Z. Yu, X. Xu, X. Chen and D. Yang In this paper, we propose a novel Convolutional Neural Network (CNN) towards cover song identification.
110 Towards Linking the Lakh and IMSLP Datasets T. Tsai We propose a method for scalable cross-modal retrieval that might be used to link the Lakh MIDI dataset with IM-SLP sheet music data.
111 A Multi-Dilation and Multi-Resolution Fully Convolutional Network for Singing Melody Extraction P. Gao, C. You and T. Chi In this paper, we propose a neural network, which includes spectro-temporal multi-resolution decomposition of the log-spectrogram of the sound and a semantic segmentation model to respectively address the bottom-up and top-down processing of hearing, for singing melody extraction.
112 An Improved Solution to the Frequency-Invariant Beamforming with Concentric Circular Microphone Arrays X. Zhao, G. Huang, J. Chen and J. Benesty In this paper, we find that the spatial aliasing problem is caused by higher-order circular harmonics.
113 Binauralaudio Source Remixing with Microphone Array Listening Devices R. M. Corey and A. C. Singer This work considers a source-remixing filter that alters the relative level of each source independently.
114 Exploiting Periodicity Features for Joint Detection and DOA Estimation of Speech Sources Using Convolutional Neural Networks R. Varzandeh, K. Adiloglu, S. Doclo and V. Hohmann In this paper, a multi-input single-output convolutional neural network (CNN) is proposed which exploits a novel feature combination for joint DOA estimation and VAD in the context of binaural hearing aids.
115 Unsupervised Multiple Source Localization Using Relative Harmonic Coefficients Y. Hu, P. N. Samarasinghe, T. D. Abhayapala and S. Gannot This paper presents an unsupervised multi-source localization algorithm using a recently introduced feature called the relative harmonic coefficients.
116 Data-Driven Wind Speed Estimation Using Multiple Microphones D. Mirabilii, K. K. Lakshminarayana, W. Mack and E. A. P. Habets A deep neural network (DNN) based approach for estimating the speed of airflows using closely-spaced microphones is proposed.
117 A Dynamic Stream Weight Backprop Kalman Filter for Audiovisual Speaker Tracking C. Schymura et al. Inspired by recent progress in the context of integrating uncertainty estimates into modern deep learning frameworks, this paper proposes a deep neural-network-based implementation of the Kalman filter with dynamic stream weights, whose parameters can be learned via standard backpropagation.
118 Maximum Likelihood Multi-Speaker Direction of Arrival Estimation Utilizing a Weighted Histogram E. Hadad and S. Gannot In this contribution, a novel maximum likelihood (ML) based direction of arrival (DOA) estimator for concurrent speakers in a noisy reverberant environment is presented.
119 Overdetermined Independent Vector Analysis R. Ikeshita, T. Nakatani and S. Araki We address the convolutive blind source separation problem for the (over-)determined case where (i) the number of nonstationary target-sources K is less than that of microphones M, and (ii) there are up to M -K stationary Gaussian noises that need not to be extracted.
120 Spatially Guided Independent Vector Analysis A. Brendel, T. Haubner and W. Kellermann We present a Maximum A Posteriori (MAP) derivation of the Independent Vector Analysis (IVA) algorithm for blind source separation incorporating an additional spatial prior over the demixing matrices.
121 Fast Independent Vector Extraction by Iterative SINR Maximization R. Scheibler and N. Ono We propose fast independent vector extraction (FIVE), a new algorithm that blindly extracts a single non-Gaussian source from a Gaussian background.
122 Regularized Fast Multichannel Nonnegative Matrix Factorization with ILRMA-Based Prior Distribution of Joint-Diagonalization Process K. Kamo et al. In this paper, we address a convolutive blind source separation (BSS) problem and propose a new extended framework of FastMNMF by introducing prior information for joint diagonalization of the spatial covariance matrix model.
123 Beyond the Dcase 2017 Challenge on Rare Sound Event Detection: A Proposal for a More Realistic Training and Test Framework J. Baumann, T. Lohrenz, A. Roy and T. Fingscheidt This paper proposes a rare SED training and test framework, which is reflecting an SED application in a more realistic way.
124 Metric Learning with Background Noise Class for Few-Shot Detection of Rare Sound Events K. Shimada, Y. Koyama and A. Inoue In this paper, we aim to achieve few-shot detection of rare sound events, from query sequence that contain not only the target events but also the other events and background noise.
125 Sound Event Detection by Multitask Learning of Sound Events and Scenes with Soft Scene Labels K. Imoto, N. Tonami, Y. Koizumi, M. Yasuda, R. Yamanishi and Y. Yamashita In this paper, we thus propose a new method for SED based on MTL of SED and ASC using the soft labels of acoustic scenes, which enable us to model the extent to which sound events and scenes are related.
126 Guided Learning for Weakly-Labeled Semi-Supervised Sound Event Detection L. Lin, X. Wang, H. Liu and Y. Qian We propose a simple but efficient method termed Guided Learning for weakly-labeled semi-supervised sound event detection (SED).
127 Staged Training Strategy and Multi-Activation for Audio Tagging with Noisy and Sparse Multi-Label Data K. He, Y. Shen, W. Zhang and J. Liu In this paper, we propose a staged training strategy to deal with the noisy label, and adopt a sigmoid-sparsemax multi-activation structure to deal with the sparse multi-label classification.
128 Learning With Out-of-Distribution Data for Audio Classification T. Iqbal, Y. Cao, Q. Kong, M. D. Plumbley and W. Wang In this paper, we investigate an instance of labelling error for classification tasks in which the dataset is corrupted with out-of-distribution (OOD) instances: data that does not belong to any of the target classes, but is labelled as such.
129 Multi-Branch Learning for Weakly-Labeled Sound Event Detection Y. Huang, X. Wang, L. Lin, H. Liu and Y. Qian Since there are only annotations for audio tagging available in weakly-supervised SED, we design multiple branches with different learning purposes instead of pursuing multiple tasks.
130 Scene-Dependent Acoustic Event Detection with Scene Conditioning and Fake-Scene-Conditioned Loss T. Komatsu, K. Imoto and M. Togami In this paper, we propose scene-dependent acoustic event detection (AED) with scene conditioning and fake-scene-conditioned loss.
131 Sound Event Localization Based on Sound Intensity Vector Refined by Dnn-Based Denoising and Source Separation M. Yasuda, Y. Koizumi, S. Saito, H. Uematsu and K. Imoto We propose a direction-of-arrival (DOA) estimation method for Sound Event Localization and Detection (SELD).
132 High-Resolution Attention Network with Acoustic Segment Model for Acoustic Scene Classification X. Bai, J. Du, J. Pan, H. Zhou, Y. Tu and C. Lee In this paper, we propose a novel strategy for acoustic scene classification, namely high-resolution attention network with acoustic segment model (HRAN-ASM).
133 Polyphonic Sound Event Detection Using Transposed Convolutional Recurrent Neural Network C. C. Chatterjee, M. Mulimani and S. G. Koolagudi In this paper we propose a Transposed Convolutional Recurrent Neural Network (TCRNN) architecture for polyphonic sound event recognition.
134 SeCoST:: Sequential Co-Supervision for Large Scale Weakly Labeled Audio Event Detection A. Kumar and V. K. Ithapu In this work, we propose a new framework for designing learning models with weak supervision by bridging ideas from sequential learning and knowledge distillation.
135 Deep Speech Extraction with Time-Varying Spatial Filtering Guided By Desired Direction Attractor Y. Nakagome, M. Togami, T. Ogawa and T. Kobayashi In this investigation, a deep neural network (DNN) based speech extraction method is proposed to enhance a speech signal propagating from the desired direction.
136 Adaptive Blind Audio Source Extraction Supervised By Dominant Speaker Identification Using X-Vectors J. Jansk?, J. M?lek, J. Cmejla, T. Kounovsk?, Z. Koldovsk? and J. ?d??nsk? We propose a novel algorithm for adaptive blind audio source extraction.
137 Convergence-Guaranteed Independent Positive Semidefinite Tensor Analysis Based on Student?s T Distribution T. Kondo et al. In this paper, we address a blind source separation (BSS) problem and propose a new extended framework of independent positive semidefinite tensor analysis (IPSDTA).
138 Determined Source Separation Using the Sparsity of Impulse Responses Y. Takahashi, D. Kitahara, K. Matsuura and A. Hirabayashi In this paper, we propose an over-determined sound source separation method considering the sparsity of impulse responses.
139 Improving Speaker Discrimination of Target Speech Extraction With Time-Domain Speakerbeam M. Delcroix et al. In this paper, we investigate strategies for improving the speaker discrimination capability of SpeakerBeam.
140 WHAMR!: Noisy and Reverberant Single-Channel Speech Separation M. Maciejewski, G. Wichern, E. McQuinn and J. L. Roux To address this, we introduce WHAMR!, an augmented version of WHAM! with synthetic reverberated sources, and provide a thorough baseline analysis of current techniques as well as novel cascaded architectures on the newly introduced conditions.
141 Impact of a Shift-Invariant Harmonic Phase Model in Fully Parametric Harmonic Voice Representation and Time/Frequency Synthesis A. Ferreira, J. Silva, F. Brito and D. Sinha In this paper, we describe two fully parametric harmonic representation and signal reconstruction alternatives that rely on a shift-invariant harmonic phase model and that implement accurate frame-based synthesis in the frequency-domain, and accurate pitch pulse-based synthesis in the time-domain.
142 Hearing aid Research Data Set for Acoustic Environment Recognition A. H?wel, K. Adiloglu and J. Bach In this work we propose a novel, binaural HA acoustic environment recognition data set (HEAR-DS) suitable for the environment recognition needs of HAs.
143 Audio Feature Extraction for Vehicle Engine Noise Classification L. Becker, A. Nelus, J. Gauer, L. Rudolph and R. Martin In this paper we propose a new scheme for vehicle engine noise classification as a more privacy-preserving alternative to classifying vehicles based on video recordings.
144 Time-Frequency Feature Decomposition Based on Sound Duration for Acoustic Scene Classification Y. Wu and T. Lee In this paper, we propose a feature decomposition method based on temporal median filtering, and use convolutional neural network to model long-duration background sounds and transient sounds separately.
145 Vggsound: A Large-Scale Audio-Visual Dataset H. Chen, W. Xie, A. Vedaldi and A. Zisserman Our goal is to collect a large-scale audio-visual dataset with low label noise from videos ?in the wild? using computer vision techniques.
146 Transfer Learning from Youtube Soundtracks to Tag Arctic Ecoacoustic Recordings E. B. ?oban, D. Pir, R. So and M. I. Mandel This paper investigates this generalization in several ways and finds that models themselves display limited performance, however, their intermediate representations can be used to train successful models on small sets of labeled data.
147 Data Augmentation Using Empirical Mode Decomposition on Neural Networks to Classify Impact Noise in Vehicle G. Nam, S. Bu, N. Park, J. Seo, H. Jo and W. Jeong In this paper, we performed data augmentation using Empirical Mode Decomposition (EMD) method that decomposes the original signal into a finite number of intrinsic mode functions (IMFs).
148 Clotho: an Audio Captioning Dataset K. Drossos, S. Lipping and T. Virtanen In this paper we present Clotho, a dataset for audio captioning consisting of 4981 audio samples of 15 to 30 seconds duration and 24 905 captions of eight to 20 words length, and a baseline method to provide initial results.
149 Robust Fundamental Frequency Estimation in Coloured Noise A. E. Jaramillo, A. Jakobsson, J. K. Nielsen and M. Gr?sb?ll Christensen To allow for this, we here propose two schemes that refine the noise statistics and parameter estimates in an iterative manner, one of them based on an approximate ML solution and the other one based on removing the periodic signal obtained from a linearly constrained minimum variance (LCMV) filter.
150 Efficient Bird Sound Detection on the Bela Embedded System A. Solomes and D. Stowell This paper proposes an automatic detection algorithm for the Bela embedded Linux device for wildlife monitoring.
151 Improving Automated Segmentation of Radio Shows with Audio Embeddings O. Berlage, K. Lux and D. Graus This study explores the novel task of using audio embeddings for automated, topically coherent segmentation of radio shows.
152 SECL-UMons Database for Sound Event Classification and Localization M. Brousmiche, J. Rouat and S. Dupont We introduce the SECL-UMons dataset for sound event classification and localization in the context of office environments.
153 Synthesizing Engaging Music Using Dynamic Models of Statistical Surprisal S. Kothinti, B. Skerritt-Davis, A. Nair and M. Elhilali In this work, we explore the statistical structure of a musical corpus and its effect on modulating the attention of listeners.
154 Harmonics Based Representation in Clarinet Tone Quality Evaluation Y. Wang, X. Guan, Y. Du and N. Nan In this paper we present a new method for identifying the clarinet reed quality by evaluating tone quality based on the harmonic structure and energy distribution.
155 Simultaneous Separation and Transcription of Mixtures with Multiple Polyphonic and Percussive Instruments E. Manilow, P. Seetharaman and B. Pardo We present a single deep learning architecture that can both separate an audio recording of a musical mixture into constituent single-instrument recordings and transcribe these instruments into a human-readable format at the same time, learning a shared musical representation for both tasks.
156 The Role of Annotation Fusion Methods in the Study of Human-Reported Emotion Experience During Music Listening T. Greer, K. Mundnich, M. Sachs and S. Narayanan In this work, we investigate several ways to represent aggregate human annotations of the complex, subjective emotional experience of listening to music.
157 Content Based Singing Voice Extraction from a Musical Mixture P. Chandna, M. Blaauw, J. Bonada and E. G?mez We present a deep learning based methodology for extracting the singing voice signal from a musical mixture based on the underlying linguistic content.
158 Neural Percussive Synthesis Parameterised by High-Level Timbral Features A. Ramires, P. Chandna, X. Favory, E. G?mez and X. Serra We present a deep neural network-based methodology for synthesising percussive sounds with control over high-level timbral characteristics of the sounds.
159 Non-Griffin?Lim Type Signal Recovery from Magnitude Spectrogram R. Nakatsu, D. Kitahara and A. Hirabayashi To make the best of such useful spectrogram transforms, we propose an algorithm which recovers the time-domain signal without the inverse spectrogram transforms.
160 Vapar Synth – A Variational Parametric Model for Audio Synthesis K. Subramani, P. Rao and A. D?Hooge We present Va-Par Synth – a Variational Parametric Synthesizer which utilizes a conditional variational autoencoder (CVAE) trained on a suitable parametric representation.
161 Bandwidth Extension of Musical Audio Signals With No Side Information Using Dilated Convolutional Neural Networks M. Lagrange and F. Gontier This paper studies the benefit of considering a dilated fully convolutional neural network to perform the bandwidth extension of musical audio signals with no side information on the magnitude spectra.
162 Towards Real-Time Single-Channel Singing-Voice Separation with Pruned Multi-Scaled Densenets M. Huber, G. Schindler, C. Sch?rkhuber, W. Roth, F. Pernkopf and H. Fr?ning We extend the multi-scaled DenseNet in several aspects to facilitate real-time source separation scenarios.
163 State-Based Transcription of Components of Carnatic Music V. S. Viraraghavan, A. Pal, H. Murthy and R. Aravind In this paper, we propose a novel state-based representation of the pitch curve motivated by CM components called constant-pitch notes and stationary points.
164 Meta-Learning Extractors for Music Source Separation D. Samuel, A. Ganeshan and J. Naradowsky We propose a hierarchical meta-learning-inspired model for music source separation (Meta-TasNet) in which a generator model is used to predict the weights of individual extractor models.
165 Consistency-Aware Multi-Channel Speech Enhancement Using Deep Neural Networks Y. Masuyama, M. Togami and T. Komatsu This paper proposes a deep neural network (DNN)?based multichannel speech enhancement system in which a DNN is trained to maximize the quality of the enhanced time-domain signal.
166 Phase Reconstruction Based On Recurrent Phase Unwrapping With Deep Neural Networks Y. Masuyama, K. Yatabe, Y. Koizumi, Y. Oikawa and N. Harada However, the training of a DNN for phase reconstruction is not an easy task because phase is sensitive to the shift of a waveform. To overcome this problem, we propose a DNN-based two-stage phase reconstruction method.
167 Performance Study of a Convolutional Time-Domain Audio Separation Network for Real-Time Speech Denoising S. Sonning, C. Sch?ldt, H. Erdogan and S. Wisdom This paper investigates the performance of such a time-domain network (Conv-TasNet) for speech denoising in a real-time setting, comparing various parameter settings.
168 Channel-Attention Dense U-Net for Multichannel Speech Enhancement B. Tolooshams, R. Giri, A. H. Song, U. Isik and A. Krishnaswamy We propose Channel-Attention Dense U-Net, in which we apply the channel-attention unit recursively on feature maps at every layer of the network, enabling the network to perform non-linear beamforming.
169 A Composite DNN Architecture for Speech Enhancement Y. Yemini, S. E. Chazan, J. Goldberger and S. Gannot In this work, we show that both separate cost functions are unsuitable for dealing with narrowband noise, and propose a new composite estimator in the log-spectrum domain.
170 Geometrically Constrained Independent Vector Analysis for Directional Speech Enhancement L. Li and K. Koishida We propose a dual-microphone speech enhancement system based on the proposed method and investigate its effectiveness with objective metrics.
171 Real-Time Speech Enhancement Using Equilibriated RNN D. Takeuchi, K. Yatabe, Y. Koizumi, Y. Oikawa and N. Harada We propose a speech enhancement method using a causal deep neural network (DNN) for real-time applications.
172 Subspace-Based Speech Correlation Vector Estimation for Single-Microphone Multi-Frame MVDR Filtering D. Fischer and S. Doclo In this paper, we propose a subspace-based estimator for the normalized speech correlation vector based on the Q largest eigenvalues and their corresponding eigenvectors of the prewhitened noisy speech correlation matrix.
173 Time-Frequency Loss for CNN Based Speech Super-Resolution H. Wang and D. Wang This paper proposes an autoencoder based fully convolutional neural network (CNN) that merges the information from both time and frequency domains.
174 Time-Domain Neural Network Approach for Speech Bandwidth Extension X. Hao, C. Xu, N. Hou, L. Xie, E. S. Chng and H. Li In this paper, we study the time-domain neural network approach for speech bandwidth extension.
175 Weighted Speech Distortion Losses for Neural-Network-Based Real-Time Speech Enhancement Y. Xia, S. Braun, C. K. A. Reddy, H. Dubey, R. Cutler and I. Tashev We propose two novel mean-squared-error-based learning objectives that enable separate control over the importance of speech distortion versus noise reduction.
176 Snorer Diarisation Based On Deep Neural Network Embeddings H. E. Romero, N. Ma and G. J. Brown This paper proposes a novel acoustic analysis system for snorer diarisation, a concept extrapolated from speaker diarisation research, which allows screening for SDB of both the user and the bed partner using a single smartphone.
177 Playing Technique Recognition by Joint Time?Frequency Scattering C. Wang, V. Lostanlen, E. Benetos and E. Chew In this paper, we propose a recognition system based on the joint time?frequency scattering transform (jTFST) for pitch evolution-based playing techniques (PETs), a group of playing techniques with monotonic pitch changes over time.
178 Privacy Aware Acoustic Scene Synthesis Using Deep Spectral Feature Inversion F. Gontier, M. Lagrange, C. Lavandier and J. Petiot Part of the research is to find ways to inform the citizen about its sound environment while ensuring her privacy.We study in this paper how this application can be cast into a feature inversion problem.
179 Robustness Assessment of Automatic Reinke?s Edema Diagnosis Systems M. Madruga, Y. Campos-Roca and C. J. P?rez The goal of this paper is to build automatic detection systems for Reinke?s edema based on a novel in-house dataset and, alternatively, on the Massachusetts Eye and Ear Infirmary Voice Disorders Database, and assess noise robustness in both cases.
180 Whosecough: In-the-Wild Cougher Verification Using Multitask Learning M. Whitehill, J. Garrison and S. Patel In this work, we overcome this problem by using multitask learning, where the second task is speaker verification.
181 Chirping up the Right Tree: Incorporating Biological Taxonomies into Deep Bioacoustic Classifiers J. Cramer, V. Lostanlen, A. Farnsworth, J. Salamon and J. P. Bello This paper introduces TaxoNet, a deep neural network for structured classification of signals from living organisms.
182 Beamforming Design for High-Resolution Low-Intensity Focused Ultrasound Neuromodulation B. Fan, W. Goodman, R. Y. Cho, S. A. Sheth, R. R. Bouchard and B. Aazhang In this study, we investigate the optimization of targeting through fine temporal and spatial power delivery control of a phased array of ultrasound elements.
183 An Attention Enhanced Multi-Task Model for Objective Speech Assessment in Real-World Environments X. Dong and D. S. Williamson In this work, we present a novel reference-less based framework called the attention enhanced multi-task speech assessment (AMSA) model, which provides reliable estimates of multiple objective quality and intelligibility measures in simulated and real-world environments.
184 Humbug Zooniverse: A Crowd-Sourced Acoustic Mosquito Dataset I. Kiskin, A. D. Cobb, L. Wang and S. Roberts We present an example use of the dataset, in which we train a convolutional neural network on log-Mel features, showcasing the information content of the labels.
185 Subjective Quality Estimation Using PESQ For Hands-Free Terminals S. Kurihara, M. Fukui, S. Shimauchi and N. Harada We propose third-party listening and conversational test procedures to assess whether PESQ can be used for predicting the subjective quality of an acoustic echo canceler.
186 Classification of Epileptic IEEG Signals by CNN and Data Augmentation X. Zhao et al. In this paper, we firstly introduce a one-dimensional convolutional neural network (1D-CNN) model for epileptic seizure focus detection which avoids the manual, time-consuming feature extraction Moreover, to reduce the necessary number of training samples, we introduce an approach for data augmentation.
187 Fractional Fourier Transform Based QRS Complex Detection in ECG Signal T. Yaqoob, S. Aziz, S. Ahmed, O. Amin and M. Alouini By exploiting fractional-Fourier-transform (FrFT), a novel technique for the QRS complex detection is proposed.
188 Cross-Domain Joint Dictionary Learning for ECG Reconstruction from PPG X. Tian, Q. Zhu, Y. Li and M. Wu This paper proposes a cross-domain joint dictionary learning (XDJDL) framework to maximize the expressive power for the two cross-domain signals.
189 An LSTM Based Architecture to Relate Speech Stimulus to Eeg M. J. Monesi, B. Accou, J. Montoya-Martinez, T. Francart and H. V. Hamme We present a novel Long Short-Term Memory (LSTM)-based architecture as a nonlinear model for the classification problem of whether a given pair of (EEG, speech envelope) correspond to each other or not.
190 Joint Semi-Supervised Feature Auto-Weighting and Classification Model for EEG-Based Cross-Subject Sleep Quality Evaluation Y. Peng, Q. Li, W. Kong, J. Zhang, B. Lu and A. Cichocki To suppress the cross-subject variances of EEG data, in this paper, we propose a joint feature auto-weighting and semi-supervised classification model, termed GRLSR, which is formulated by introducing an auto-weighting variable into the least square regression to adaptively and quantitatively measure the importance of each dimension of the feature.
191 Reversal No Longer Matters: Attention-Based Arrhythmia Detection with Lead-Reversal ECG Data Z. Cao, J. Shi and J. Wu In this paper, we propose an attention-based multi-scale neural network for arrhythmia detection with lead-reversal electrocardiogram data.
192 Augmenting Molecular Images with Vector Representations as a Featurization Technique for Drug Classification D. d. Marchi and A. Budhiraja This paper proposes the creation of molecular images “captioned” with binary vectors that encode information not contained in or easily understood from a molecular image alone.
193 Multi-Modal Self-Supervised Pre-Training for Joint Optic Disc and Cup Segmentation in Eye Fundus Images ?. S. Hervella, L. Ramos, J. Rouco, J. Novo and M. Ortega This paper presents a novel approach for the segmentation of the optic disc and cup in eye fundus images using deep learning.
194 Dense Mapping of Intracellular Diffusion and Drift from Single-Particle Tracking Data Analysis A. Salomon, C. A. Valades-Cruz, L. Leconte and C. Kervrann In this paper, we propose a new mapping method to robustly estimate dynamics in the entire cell from particle tracks.
195 A Deep Gradient Boosting Network for Optic Disc and Cup Segmentation Q. Liu, B. Zou, Y. Zhao and Y. Liang To build connections among prediction branches, this paper introduces gradient boosting framework to deep classification model and proposes a gradient boosting network called BoostNet.
196 Adaptive Elastic Loss Based on Progressive Inter-Class Association for Cervical Histology Image Segmentation Z. Meng, Z. Zhao, F. Su and W. Wang In this paper, an end-to-end deep segmentation network for complex cervical histology images is proposed, and a benchmark evaluation is contributed.
197 A Bidirectional Context Propagation Network for Urine Sediment Particle Detection in Microscopic Images M. Yan, Q. Liu, Z. Yin, D. Wang and Y. Liang This paper proposes a bidirectional context propagation network called BCPNet for urine sediment particle detection.
198 The Swax Benchmark: Attacking Biometric Systems with Wax Figures R. H. Vareto, A. Marcia Saldanha and W. R. Schwartz This work introduces a new database named Sense Wax Attack dataset (SWAX), comprised of real human and wax figure images and videos that endorse the problem of face spoofing detection.
199 Resting-State EEG-Based Biometrics with Signals Features Extracted by Multivariate Empirical Mode Decomposition M. K. Ma, T. Lee, M. C. Fong and W. Shiyuan Wang This study highlights the uniqueness of the proposed non-stationary and connectivity-based feature and demonstrated its success as a biometrics.
200 Auto-Fas: Searching Lightweight Networks for Face Anti-Spoofing Z. Yu et al. In this paper, we propose a neural architecture search (NAS) based method called Auto-FAS, intending to discover well-suitable lightweight networks for mobile-level face anti-spoofing.
201 Domain Adaptation for Generalization of Face Presentation Attack Detection in Mobile Settengs with Minimal Information A. Mohammadi, S. Bhattacharjee and S. Marcel Here, we propose a novel one class domain adaptation method which uses domain guided pruning to adapt a pre-trained PAD network to the target dataset.
202 A Lightweight Multi-Label Segmentation Network for Mobile Iris Biometrics C. Wang, Y. Wang, B. Xu, Y. He, Z. Dong and Z. Sun This paper proposes a novel, lightweight deep convolutional neural network specifically designed for iris segmentation of noisy images acquired by mobile devices.
203 Modeling Behavioral Consistency in Large-Scale Wearable Recordings of Human Bio-Behavioral Signals T. Feng and S. S. Narayanan In this work, we introduce a novel data processing pipeline to model behavioral consistency in a large real-world wearable recording data-set collected in a hospital workplace setting from nurses and direct clinical providers for a period of ten weeks.
204 Modeling Behavior as Mutual Dependency between Physiological Signals and Indoor Location in Large-Scale Wearable Sensor Study T. Feng, B. M. Booth and S. S. Narayanan In this work, we describe our exploration in discovering the correlation between one?s physiological responses and movement patterns within different indoor locations using data collected from nurses in a hospital workplace for a ten week period.
205 Multichannel Signal Classification Using Vector Autoregression A. Haboub, H. Baali and A. Bouzerdoum To further explore this area, we propose a simple yet effective approach based on modeling MCS with a VAR process.
206 Efficient Algorithm to Implement Sliding Singular Spectrum Analysis with Application to Biomedical Signal Denoising M. Saeed, C. C. Took and S. R. Alty In this paper, we show it is possible to evaluate the rank-1 SVD update efficiently in $\mathcal{O}\left( {{n^2}} \right)$, thus dramatically increasing the speed of the sliding version of the SSA algorithm.
207 Strategic Attention Learning for Modality Translation J. Martinez, A. Akbari, K. Sel and R. Jafari In this work, we propose a two-stage deep learning framework that leverages a novel attention mechanism to translate Bio-Z signals to highly interpretable electrocardiogram (ECG) waveforms while also predicting translation uncertainty.
208 Sparse CSP Algorithm via Joint Spatio-Temporal Filtering A. Jiang, J. Shang, W. Cheng, X. Liu, H. K. Kwan and Y. Zhu To improve its performance, a novel and efficient spatio-temporal filtering strategy is proposed in this paper to extract discriminative features.
209 Human-Machine Collaboration for Medical Image Segmentation M. Ravanbakhsh et al. In this paper, we propose a method based on conditional Generative Adversarial Network (cGAN) to address segmentation in semi-supervised setup and in a human-in-the-loop fashion.
210 Mixup Multi-Attention Multi-Tasking Model for Early-Stage Leukemia Identification P. Mathur, M. Piplani, R. Sawhney, A. Jindal and R. R. Shah To this effect, we propose a novel architecture termed as Mixup Multi-Attention Multi-Task Learning Model (MMA-MTL), which introduces Pointwise Attention Convolution Layers and Local Spatial Attention blocks to capture global and local features simultaneously.
211 Cross-View Attention Network for Breast Cancer Screening from Multi-View Mammograms X. Zhao, L. Yu and X. Wang In this paper, we address the problem of breast caner detection from multi-view mammograms.
212 UNet 3+: A Full-Scale Connected UNet for Medical Image Segmentation H. Huang et al. In this paper, we propose a novel UNet 3+, which takes advantage of full-scale skip connections and deep supervisions.
213 Unsupervised Content-Preserved Adaptation Network for Classification of Pulmonary Textures from Different CT Scanners R. Xu, Z. Cong, X. Ye, S. Kido and N. Tomiyama In this paper, we propose an unsupervised content-preserved adaptation network to address this problem.
214 Classify and Explain: An Interpretable Convolutional Neural Network For Lung Cancer Diagnosis Y. Li, D. Gu, Z. Wen, F. Jiang and S. Liu In this paper, we present a novel network structure for visually interpretable lung nodule diagnosis.
215 Robust Global Optimized Affine Registration Method for Microscopic Images of Biological Tissue Y. Lv, X. Chen, C. Shu and H. Han In this paper, a global optimized affine registration method is proposed, which can be used in volume reconstruction.
216 Empirical Sure-Guided Microscopy Super-Resolution Image Reconstruction from Confocal Multi-Array Detectors S. Prigent, S. Dutertre and C. Kervrann In this paper, we review the most commonly used reconstruction methods and propose a SURE approach to automatically estimate parameters and improve reconstruction.
217 Encoding Temporal Information For Automatic Depression Recognition From Facial Analysis W. C. de Meto, E. Granger and M. B. Lopez To address these issues, we propose a novel temporal pooling method to capture and encode the spatio-temporal dynamic of video clips into an image map.
218 Retinal Vessel Segmentation via a Semantics and Multi-Scale Aggregation Network R. Xu, X. Ye, G. Jiang, T. Liu, L. Li and S. Tanaka In this paper, we propose a semantics and multi-scale aggregation network to address these difficulties.
219 Adaptive Matched Filter using Non-Target Free Training Data A. M. Rekavandi, A. Seghouane and R. J. Evans To develop the test, an ad hoc approach, similar to the classical adaptive matched filter (AMF) is used where instead of the maximum likelihood (ML) estimator of the covariance, the minimum a-divergence based estimator is substituted in the likelihood ratio.
220 Feature Drift Resilient Tracking of The Carotid Artery Wall Using Unscented Kalman Filtering With Data Fusion J. Dorazil, R. Repp, T. Kropfreiter, R. Pr?ller, K. R?ha and F. Hlawatsch Here, we propose a method for tracking CCA wall motion from a B-mode ultrasound video sequence.
221 Tracing Network Evolution Using The Parafac2 Model M. Roald, S. Bhinge, C. Jia, V. Calhoun, T. Adali and E. Acar In this paper, without assuming static networks in time and/or space, we arrange the temporal data as a higher-order tensor and use a tensor fac-torization model called PARAFAC2 to capture underlying patterns (spatial networks) in time-evolving data and their evolution.
222 A Model-Based Deep Network for MRI Reconstruction Using Approximate Message Passing Algorithm X. Qiao, J. Du, L. Wang, Z. He and Y. Jia We propose a novel model-based network to reconstruct the magnetic resonance (MR) image.
223 Online Positron Emission Tomography By Online Portfolio Selection Y. Li The number of measurement outcomes in positron emission tomography (PET) is typically large, rendering signal reconstruction computationally expensive. We propose an online algorithm to address this computational issue.
224 Space Filling Curves for MRI Sampling S. Sharma, K. V. S. Hari and G. Leus We propose 1-shot and 4-shot variable density SFCs by utilizing the space coverage provided by SFCs in different iterations.
225 K-Space Trajectory Design for Reduced MRI Scan Time S. Sharma, K. V. S. Hari and G. Leus In this paper, a constrained convex optimization based method to obtain feasible trajectories is proposed.
226 Rethinking Retinal Landmark Localization as Pose Estimation: Na?ve Single Stacked Network for Optic Disk and Fovea Detection S. R. Maiya and P. Mathur In this regard, we present Naive Single Stacked Hourglass (NSSH) network which learns the spatial orientation and pixel intensity contrast between optic disk and fovea to accurately pinpoint their locations.
227 Hidden Markov Models for Sepsis Detection in Preterm Infants A. Honor? et al. To improve the neural network based HMM, we propose a discriminative training approach.
228 Blood Pressure Estimation From PPG Signals Using Convolutional Neural Networks And Siamese Network O. Schlesinger, N. Vigderhouse, D. Eytan and Y. Moshe This paper presents two techniques that enable continuous and noninvasive cuff-less BP estimation using photoplethysmography (PPG) signals with Convolutional Neural Networks (CNNs).
229 Speech Breathing Estimation Using Deep Learning Methods V. S. Nallanthighal, A. H?rm? and H. Strik In this paper, we explore techniques for sensing breathing from speech using deep learning architectures including multi-task learning approaches.
230 A Fast Non-Contact Vital Signs Detection Method Based on Regional Hidden Markov Model in A 77ghz Lfmcw Radar System Z. Mei, Q. Wu, Z. Hu and J. Tao In this paper, a 77GHz linear frequency modulated continuous-wave (LFMCW) radar system is investigated to mitigate multiple-targets interferences.
231 Robust Likelihood Ratio Test Using a-Divergence A. M. Rekavandi, A. Seghouane and R. J. Evans The problem of detecting a subspace signal in the presence of subspace interference and contaminated Gaussian noise with unknown variance is investigated.
232 Using X-Vectors to Automatically Detect Parkinson?s Disease from Speech L. Moro-Velazquez, J. Villalba and N. Dehak Accordingly, in this study we analyze for the first time a new state-of-the-art speaker recognition technique, x-Vectors, in a different scenario: the automatic detection of PD from speech.
233 Learning A Common Granger Causality Network Using A Non-Convex Regularization P. Manomaisaowapak and J. Songsiri This paper proposes an estimation for learning a common Granger network of panel data.
234 Synthetic Data Generation Through Statistical Explosion: Improving Classification Accuracy of Coronary Artery Disease Using PPG S. Bhattacharya, O. Mazumder, D. Roy, A. Sinha and A. Ghose This paper presents a novel approach of generating synthetic Photoplethysmogram (PPG) data using statistical explosion.
235 High-Accuracy Classification of Attention Deficit Hyperactivity Disorder with L2,1-Norm Linear Discriminant Analysis Y. Tang, X. Li, Y. Chen, Y. Zhong, A. Jiang and X. Liu In detail, we introduce a binary hypothesis testing framework as the classification outline to cope with insufficient data of ADHD database.
236 A Neural Network-Based Spike Sorting Feature Map That Resolves Spike Overlap in the Feature Space J. Wouters, F. Kloosterman and A. Bertrand In this work, a novel approach is presented to resolve spike overlap directly in the feature space.
237 Gait Phase Segmentation Using Weighted Dynamic Time Warping and K-Nearest Neighbors Graph Embedding T. Chen, T. Lin and Y. -. P. Hong To reduce the complexity of the DTW-based kNN search, we propose a neural network-based graph embedding scheme that is able to map the IMU signals associated with each gait cycle into a distance-preserving low-dimensional representation while also producing a prediction on the k nearest neighbors of the test signal.
238 Automatic Classification of Volumes of Water Using Swallow Sounds from Cervical Auscultation S. Subramani, M. V. Achuth Rao, D. Giridhar, P. S. Hegde and P. Kumar Ghosh In this work, a rich set of acoustic features, the ComParE 2016 acoustic feature set (OS) is used to investigate whether several temporal, spectral, vocal and source features and their functionals provide cues for volume classification.
239 Conditional Domain Adversarial Transfer for Robust Cross-Site ADHD Classification Using Functional MRI Y. Huang, W. Hsieh, H. Yang and C. Lee Hence in this research, we utilize an approach of conditional adversarial domain adaptation network (CDAN) to learn a discriminative fMRI representation that is site-invariant for unsupervised transfer of ADHD classification.
240 Eeg Connectivity – Informed Cooperative Adaptive Line Enhancer for Recognition of Brain State S. Sanei, C. C. Took, D. Jarchi and A. Prochazka In this paper, it is shown that for a proposed cooperative adaptive line enhancer, which can both detect and separate such periodic bursts, the combination weights are consistently different from each other.
241 Online Graph Topology Inference with Kernels For Brain Connectivity Estimation M. Moscu, R. Borsoi and C. Richard This paper focuses on estimating in an online and adaptive manner a network structure capturing the non-linear dependencies among streaming graph signals in the form of a possibly directed, adjacency matrix.
242 Minimal Adversarial Perturbations in Mobile Health Applications: The Epileptic Brain Activity Case Study A. Aminifar In this article, we demonstrate the power of such adversarial attacks based on a real-world epileptic seizure detection problem.
243 Detecting Autism Spectrum Disorder Using Topological Data Analysis S. Majumder, F. Apicella, F. Muratori and K. Das We used a feature extraction technique based on Topological Data Analysis (TDA) to classify autistic subjects from typically developing ones.
244 Multi-View Bayesian Generative Model for Multi-Subject FMRI Data on Brain Decoding of Viewed Image Categories Y. Akamatsu, R. Harakawa, T. Ogawa and M. Haseyama In this paper, we propose a multi-view Bayesian generative model for multi-subject fMRI data to estimate viewed image categories from fMRI activity.
245 Time-Frequency Analysis of Unimodal Sensory Processing In Autism Spectrum Disorder D. F. D?Croz-Baron, M. C. Baker and T. Karp This work summarizes the results of a time-frequency analysis of sensory processing in young adults with Autism Spectrum Disorder via continuous wavelet transform.
246 Automatic Epileptic Seizure Onset-Offset Detection Based On CNN in Scalp EEG P. Boonyakitanont, A. Lek-uthai and J. Songsiri We establish a deep learning-based method to automatically detect the epileptic seizure onsets and offsets in multi-channel electroencephalography (EEG) signals.
247 Enhance feature representation of electroencephalogram for Seizure detection D. Wang, Y. Fang, Y. Li and C. Chai In order to solve these problems, we propose a novel EEG feature pre-presentation method for seizure detection based on the Log Mel-Filterbank energy feature.
248 Speech Synthesis Using EEG G. Krishna, C. Tran, Y. Han, M. Carnahan and A. H. Tewfik In this paper we demonstrate speech synthesis using different electroencephalography (EEG) feature sets recently introduced in [1].
249 Eeg Feature Selection Using Orthogonal Regression: Application to Emotion Recognition X. Xu, F. Wei, Z. Zhu, J. Liu and X. Wu To reduce the redundancy and choose informative EEG features, in this paper, we propose an EEG feature selection technique, termed as Feature Selection with Orthogonal Regression (FSOR).
250 Scalpnet: Detection of Spatiotemporal Abnormal Intervals in Epileptic EEG Using Convolutional Neural Networks T. Sakai, T. Shoji, N. Yoshida, K. Fukumori, Y. Tanaka and T. Tanaka We propose ScalpNet: A deep neural network to detect spatiotemporal abnormal intervals from EEGs of epilepsy patients.
251 A Semi-Supervised Approach For Identifying Abnormal Heart Sounds Using Variational Autoencoder R. Banerjee and A. Ghose In this paper, we propose a semi-supervised approach to solve the problem.
252 Detection Of S1 And S2 Locations In Phonocardiogram Signals Using Zero Frequency Filter R. Prasad, G. Yilmaz, O. Chetelat and M. Magimai.-Doss In this paper, we propose a purely time?domain processing based method that employs a heavily decaying low pass filter (referred to as zero frequency filter) to suppress extraneous factors and detect S1?S2 locations.
253 Mental Fatigue Prediction from Multi-Channel ECOG Signal L. Yao, J. L. Baker, J. Ryou, N. D. Schiff, K. P. Purpura and M. Shoaran In this study, we analyzed electrocorticography (ECoG) signals chronically recorded from two non-human primates (NHPs) as they performed a cognitively demanding task over extended periods of time.
254 The Effect of Data Augmentation on Classification of Atrial Fibrillation in Short Single-Lead ECG Signals Using Deep Neural Networks F. N. Hatamian, N. Ravikumar, S. Vesal, F. P. Kemeth, M. Struck and A. Maier In this study, we investigate the impact of various data augmentation algorithms, e.g., oversampling, Gaussian Mixture Models (GMMs) and Generative Adversarial Networks (GANs), on solving the class imbalance problem.
255 Atrial Fibrillation Risk Prediction from Electrocardiogram and Related Health Data with Deep Neural Network Y. Chen, A. H. Twing, D. Badawi, J. Danavi, M. McCauley and A. E. Cetin In this study, we develop a novel and effective method to predict the potential AF risk of patients using our ECG signal dataset collected in the University of Illinois Hospital and Health Sciences System.
256 Adaptive Region Aggregation Network: Unsupervised Domain Adaptation with Adversarial Training for ECG Delineation M. Chen, G. Wang, H. Chen and Z. Ding In this work, we propose an unsupervised domain adaptation method called Adaptive Region Aggregation Network (ARAN) based on adversarial training to tackle domain shift problem in ECG delineation.
257 Matching Pursuit Based Dynamic Phase-Amplitude Coupling Measure T. T. K. Munia and S. Aviyente In this paper, we introduce a data-driven approach to quantify dynamic PAC.
258 Multitaper Spectral Granger Causality with Application to Ssvep R. Anderson and M. Sandsten These limits can be solved by using nonparametric spectral estimates in the frequency-domain formulation of GC, also known as spectral GC.
259 Cross-Domain Adaptation for Biometric Identification Using Photoplethysmogram E. Lee, A. Ho, Y. Wang, C. Huang and C. Lee In this work, we propose the use of both unsupervised and semi-supervised adversarial learning techniques for cross-domain adaptation.
260 Exploring Bio-Behavioral Signal Trajectories of State Anxiety During Public Speaking E. H. Nirjhar, A. Behzadan and T. Chaspari We tackle this problem by introducing temporal parametric models to quantify bio-behavioral trajectories of PSA throughout a public speaking encounter.
261 Real-Time Hand Gesture Recognition Using Temporal Muscle Activation Maps of Multi-Channel Semg Signals A. D. Silva, M. V. Perera, K. Wickramasinghe, A. M. Naim, T. Dulantha Lalitharatne and S. L. Kappel Based on these maps, we propose an algorithm that can recognize hand gestures in real-time using a Convolution Neural Network.
262 XceptionTime: Independent Time-Window Xceptiontime Architecture for Hand Gesture Classification E. Rahimian, S. Zabihi, S. F. Atashzar, A. Asif and A. Mohammadi Capitalizing on the goal of addressing identified shortcomings of recent solutions developed for recognition tasks via sparse multichannel surface Electromyography (sEMG) signals, the paper proposes a novel deep learning model, referred to as the XceptionTime architecture.
263 Discovering Causalities from Cardiotocography Signals using Improved Convergent Cross Mapping with Gaussian Processes G. Feng, J. G. Quirk and P. M. Djuric In this paper, we propose a novel improved version of CCM using Gaussian processes for discovery of causality from noisy time series.
264 Lai-Net: Local-Ancestry Inference with Neural Networks D. M. Montserrat, C. Bustamante and A. Ioannidis In this paper we develop the first neural network based LAI method, named LAI-Net, providing competitive accuracy with state-of-the-art methods and robustness to missing or noisy data, while having a small number of layers.
265 Prediction of Individual Progression Rate in Parkinson?s Disease Using Clinical Measures and Biomechanical Measures of Gait and Postural Stability V. Raval, K. P. Nguyen, A. Gerald, R. B. Dewey and A. Montillo The primary aim of this study was to develop a model using clinical measures and biomechanical measures of gait and postural stability to predict an individual?s PD progression over two years.
266 Deep Matrix Completion on Graphs: Application in Drug Target Interaction Prediction A. Mongia and A. Majumdar This work proposes matrix completion via deep matrix factorization on graphs.
267 Identification of Essential Proteins Using A Novel Multi-Objective Optimization Method C. Wu, H. Zhang, L. Zhang and H. Zheng Hence, in this paper, we consider the identification of essential proteins as a multi-objective optimization problem and use a novel multi-objective optimization method to solve it.
268 Graph Convolutional Neural Networks to Classify Whole Slide Images R. Konda, H. Wu and M. D. Wang Here, we present a novel application of graph convolutional networks (GCNs) to analyze WSIs.
269 Deep James-Stein Neural Networks For Brain-Computer Interfaces M. Angjelichinoski, M. Soltani, J. Choi, B. Pesaran and V. Tarokh We propose a novel method that combines James-Stein regression for feature extraction, and deep neural network for decoding; we refer to the architecture as deep James-Stein neural network (DJSNN).
270 Formulating Divergence Framework for Multiclass Motor Imagery EEG Brain Computer Interface S. Kumar, T. K. Reddy, V. Arora and L. Behera In this work, a novel method is proposed based on Joint Approximate Diagonalization (JAD) to optimize stationarity for multiclass motor imagery Brain Computer Interface (BCI) in an information theoretic framework.
271 Subject Transfer Framework Based on Source Selection and Semi-Supervised Style Transfer Mapping for Semg Pattern Recognition S. Kanoga, T. Hoshino and H. Asoh In this study, we investigate the efficiency of the proposed subject transfer framework, which applies a discriminability-based source selection approach and semi-supervised style transfer mapping algorithm, by constructing support vector machine classifiers.
272 Decoding Movement Imagination and Execution From Eeg Signals Using Bci-Transfer Learning Method Based on Relation Network D. Lee, J. Jeong, K. Shim and S. Lee In this study, we focused on a robust MI decoding method with transfer learning for the ME and MI paradigm.
273 Classification of High-Dimensional Motor Imagery Tasks Based on An End-To-End Role Assigned Convolutional Neural Network B. Lee, J. Jeong, K. Shim and S. Lee In this study, we collected intuitive EEG data contained the nine different types of movements of a single-arm from 9 subjects. We propose an end-to-end role assigned convolutional neural network (ERA-CNN) which considers discriminative features of each upper limb region by adopting the principle of a hierarchical CNN architecture.
274 Channel Selection over Riemannian Manifold with Non-Stationarity Consideration for Brain-Computer Interface Applications K. Sadatnejad, A. Roc, L. Pillette, A. appriou, T. Monseigne and F. Lotte In this paper, we propose and compare multiple criteria for selecting ElectroEncephaloGraphic (EEG) channels over the Riemannian manifold, for EEG classification in Brain- Computer Interfaces (BCI).
275 A Segmentation Based Robust Deep Learning Framework for Multimodal Retinal Image Registration Y. Wang et al. In this paper, a deep learning framework for multimodal retinal image registration is proposed.
276 Dense Residual Network for Retinal Vessel Segmentation C. Guo, M. Szemenyei, Y. Yi, Y. Xue, W. Zhou and Y. Li In this work, we propose an efficient method to segment blood vessels in Scanning Laser Ophthalmoscopy (SLO) retinal images.
277 Lightweight V-Net for Liver Segmentation T. Lei, W. Zhou, Y. Zhang, R. Wang, H. Meng and A. K. Nandi To address these issues, we design a lightweight V-Net (LV-Net) for liver segmentation in this paper.
278 Acu-Net: A 3D Attention Context U-Net for Multiple Sclerosis Lesion Segmentation C. hu, G. Kang, B. Hou, Y. Ma, F. Labeau and Z. Su To solve the problem, 3D attention context U-Net (ACU-Net) is proposed for MS lesion segmentation in this paper.
279 EDNFC-Net: Convolutional Neural Network with Nested Feature Concatenation for Nuclei-Instance Segmentation S. Gehlot, A. Gupta and R. Gupta To address these challenges, we propose an Encoder-Decoder based Convolutional Neural Network (CNN) with Nested-Feature Concatenation (EDNFC-Net) for automatic nuclei segmentation.
280 An Unsupervised Retinal Vessel Extraction and Segmentation Method Based On a Tube Marked Point Process Model T. Li, M. Comer and J. Zerubia In this paper, we propose an unsupervised segmentation method based on our previous connected tube marked point process (MPP) model.
281 KALM: Key Area Localization Mechanism for Abnormality Detection in Musculoskeletal Radiographs W. Huang, Z. Xiong, Q. Wang and X. Li To achieve this goal, we propose a key area localization mechanism (KALM) for abnormality detection for the first time in this paper.
282 Combining CGAN and Mil for Hotspot Segmentation in Bone Scintigraphy H. Xu, S. Geng, Y. Qiao, K. Xu and Y. Gu In this paper, we propose a new framework to detect and extract hotspots in thoracic region by integrating the techniques of both conditional generative adversarial networks (cGAN) and multiple instance learning (MIL).
283 A Noninvasive Method to Detect Diabetes Mellitus and Lung Cancer Using the Stacked Sparse autoencoder Q. Zhang, J. Zhou and B. Zhang The aim of this paper is to distinguish patients with diabetes mellitus, lung cancer from healthy people simultaneously by analyzing facial images through the stacked sparse autoencoder.
284 A Multi-Scaled Receptive Field Learning Approach for Medical Image Segmentation P. Guo, X. Su, H. Zhang, M. Wang and F. Bao To solve this problem, this paper integrates an atrous spatial pyramid pooling (ASPP) module in the contracting path of attention U-Net.
285 Automatic Data Augmentation Via Deep Reinforcement Learning for Effective Kidney Tumor Segmentation T. Qin, Z. Wang, K. He, Y. Shi, Y. Gao and D. Shen In this paper, we developed a novel automatic learning-based data augmentation method for medical image segmentation which models the augmentation task as a trial-and-error procedure using deep reinforcement learning (DRL).
286 Cross-Stained Segmentation from Renal Biopsy Images Using Multi-Level Adversarial Learning K. Mei, C. Zhu, L. Jiang, J. Liu and Y. Qiao In this paper, we design a robust and flexible model for cross-stained segmentation.
287 Blind Multi-Spectral Image Pan-Sharpening L. Yu, D. Liu, H. Mansour, P. T. Boufounos and Y. Ma We address the problem of sharpening low spatial-resolution multi-spectral (MS) images with their associated misaligned high spatial-resolution panchromatic (PAN) image, based on priors on the spatial blur kernel and on the cross-channel relationship.
288 A Forward-Backward Algorithm for Reweighted Procedures: Application to Radio-Astronomical Imaging A. Repetti and Y. Wiaux In this work we present a reweighted forward-backward algorithm designed to handle non-convex composite functions.
289 Cra: A Generic Compression Ratio Adapter for End-To-End Data-Driven Image Compressive Sensing Reconstruction Frameworks Z. Zhang, K. Xu and F. Ren This paper presents a generic compression ratio adapter (CRA) framework that addresses the variable compression ratio (CR) problem for existing EDCSR frameworks with no modification to given reconstruction models nor enormous rounds of training needed.
290 Revealing Hidden Drawings in Leonardo?s ?the Virgin of the Rocks? from Macro X-Ray Fluorescence Scanning Data through Element Line Localisation S. Yan, J. Huang, N. Daly, C. Higgitt and P. L. Dragotti In this paper, we propose a new method that can process macro XRF scanning data from paintings fully automatically.
291 3D Unknown View Tomography Via Rotation Invariants M. Zehni, S. Huang, I. Dokmanic and Z. Zhao In this paper, we study the problem of reconstructing a 3D point source model from a set of 2D projections at unknown view angles.
292 Modelling Sea Clutter In Sar Images Using Laplace-Rician Distribution O. Karakus, E. E. Kuruoglu and A. Achim This paper presents a novel statistical model for the characterisation of synthetic aperture radar (SAR) images of the sea surface.
293 Volume Reconstruction for Light Field Microscopy H. Verinaz-Jadan, P. Song, C. L. Howe, A. J. Foust and P. L. Dragotti In this work, we make two contributions. First, we propose a simplification of the forward model based on a novel discretization approach that allows us to accelerate the computation without drastically increasing memory consumption. Second, we experimentally show that by including regularization priors and an appropriate initialization strategy, it is possible to remove the artifacts near the native object plane.
294 Deep Exposure Fusion with Deghosting via Homography Estimation and Attention Learning S. Chen and Y. Chuang This paper proposes a deep network for exposure fusion.
295 Single-Shot Real-Time Multiple-Path Time-of-Flight Depth Imaging for Multi-Aperture and Macro-Pixel Sensors M. H. Conde, K. Kagawa, T. Kokado, S. Kawahito and O. Loffeld In this work we consider two hardware alternatives able to acquire all necessary raw data in a single shot.
296 Fast Optical System Identification by Numerical Interferometry S. Gupta, R. Gribonval, L. Daudet and I. Dokmanic We propose a numerical interferometry method for identification of optical multiply-scattering systems when only intensity can be measured.
297 Fourier Phase Retrieval with Arbitrary Reference Signal F. Arab and M. S. Asif In this paper, we present a Fourier phase retrieval algorithm in the presence of a known (reference) signal at arbitrary location in the scene.
298 Preconditioned Ghost Imaging Via Sparsity Constraint Z. Tong, J. Wang and S. Han To deal with this issue, we propose an efficient recovery algorithm for GISC called the preconditioned multiple orthogonal least squares (PmOLS).
299 Multispectral Fusion of RGB and NIR Images Using Weighted Least Squares and Alternating Guidance K. Zhou and C. Jung In this paper, we propose multi-spectral fusion of RGB and NIR images using weighted least squares (WLS) and alternating guidance.
300 Color and Angular Reconstruction of Light Fields from Incomplete-Color Coded Projections H. Nguyen and C. Guillemot We present a simple variational approach for reconstructing color light fields (LFs) in the compressed sensing (CS) framework with very low sampling ratio, using both coded masks and color filter arrays (CFAs).
301 Cross Image Cubic Interpolator for Spatially Varying Exposures Z. Li, J. Zheng, S. Xie and H. Shu In this paper, we introduce a novel cross image cubic interpolator for the spatially varying exposures via the rolling shutter.
302 Discriminant and Sparsity Based Least Squares Regression with l1 Regularization for Feature Representation S. Zhao, B. Zhang and S. Li To solve these dilemmas, this paper presents a discriminant and sparsity based least squares regression with l1 regularization (DS_LSR).
303 Deep Meta-Relation Network for Visual Few-Shot Learning F. Zhang, Q. Wang and X. Li This paper proposes a novel metric-based deep learning method to solve the few-shot learning problem.
304 A Semi-Supervised Rank Tracking Algorithm For On-Line Unmixing Of Hyperspectral Images L. NUS, S. MIRON, B. JAILLAIS, S. MOUSSAOUI and D. BRIE Based on the On-line Alternating Direction Method of Multipliers (ADMM), we propose a new hyperspectral unmixing approach that integrates prior information as well as joint sparsity regularization, allowing to select only the active components on each sample of the image.
305 Inverse Multiple Scattering with Phaseless Measurements M. A. Lodhi, Y. Ma, H. Mansour, P. T. Boufounos and D. Liu We study the problem of reconstructing an object from phaseless measurements in the context of inverse multiple scattering.
306 Multi-Polarization Information Fusion for Object Contour Display in Passive Millimeter-Wave and Terahertz Security Imaging Y. Cheng, Z. Zhao, Y. Wang and Y. Niu In this paper, a physical-based contour display method by fusing multi-polarization information is proposed.
307 Characterisation of a Snapshot Fourier Transform Imaging Spectrometer Based on an Array of Fabry-Perot Interferometers D. Picone, A. Dolet, S. Gousset, D. Voisin, M. D. Mura and E. Le Coarer In this paper, we present a strategy for estimating the thickness of the Fabry-Perot cavities, as this information is typically not precise or even available.
308 Shadow Removal of Text Document Images by Estimating Local and Global Background Colors J. Wang and Y. Chuang This paper proposes a simple yet effective method for removing shadows from text document images.
309 Exploring Energy Efficient Quantum-resistant Signal Processing Using Array Processors H. Nejatollahi, S. Shahhosseini, R. Cammarota and N. Dutt In this work, we explore the energy efficiency of polynomial multiplier using systolic architecture for the first time.
310 dMazeRunner: Optimizing Convolutions on Dataflow Accelerators S. Dave, A. Shrivastava, Y. Kim, S. Avancha and K. Lee We propose dMazeRunner framework, which allows users to optimize execution methods for accelerating convolution and matrix multiplication on a given architecture and to explore dataflow accelerator designs for efficiently executing CNN models.
311 Exploration Methodology for BTI-Induced Failures on RRAM-Based Edge AI Systems A. Levisse, M. Rios, M. Pe?n-Quir?s and D. Atienza Based on this observation, in this work, we propose to explore how Edge-level applications running on a RRAM-based Edge device could fail because of Bias Temperature Instability (BTI).
312 Time-Predictable Software-Defined Architecture with Sdf-Based Compiler Flow for 5g Baseband Processing V. Venkataramani, B. Bodin, A. Kulkarni, T. Mitra and L. Peh We introduce a software-defined array-based many-core architecture, called SPECTRUM, that couples lightweight predictable hardware components with a compiler flow that orchestrates the on-chip hardware resources.
313 Accelerating Linear Algebra Kernels on a Massively Parallel Reconfigurable Architecture A. Soorishetty et al. We present customized implementation of a select set of linear algebra kernels, namely, triangular matrix solver, LU decomposition, QR decomposition and matrix in-version, on Transformer.
314 Energy Efficient Acceleration Of Floating Point Applications Onto CGRA S. Das, R. Prasad, K. J. M. Martin and P. Coussy In this paper, we propose a novel CGRA architecture and associated compilation flow supporting both integer and floating-point computations for energy efficient acceleration of DSP applications.
315 Fast and Accurate Embedded DCNN for Rgb-D Based Sign Language Recognition C. Wang, C. Chiu, C. Huang, Y. Ding and L. Wang In this paper, fast and accurate two paths CNN architecture was designed in hardware-oriented manner.
316 D2NA: Day-To-Night Adaptation for Vision based Parking Management System W. Zheng, V. Tran and C. Huang In this paper, we propose a novel framework based on day-night domain adaptation, feature disentanglement, and style transfer to transfer the knowledge from day to night.
317 EnerGAN: A GENERATIVE ADVERSARIAL NETWORK FOR ENERGY DISAGGREGATION M. Kaselimi, A. Voulodimos, E. Protopapadakis, N. Doulamis and A. Doulamis An efficient, appliance-level approach for energy disaggregation, exploiting the benefits of Generative Adversarial Networks, is presented.
318 Enhancing the Labelling of Audio Samples for Automatic Instrument Classification Based on Neural Networks G. Castel-Branco, G. Falcao and F. Perdig?o In the paper we disambiguate and report the automatic labelling of previously unlabelled samples.
319 Deep-Neural-Network Based Fall-Back Mechanism in Interference-Aware Receiver Design S. Hu, W. Hu, D. Kapetanovic and N. Wang In this paper, we consider designing a fall-back mechanism in an interference-aware receiver.
320 DNN-Chip Predictor: An Analytical Performance Predictor for DNN Accelerators with Various Dataflows and Hardware Architectures Y. Zhao, C. Li, Y. Wang, P. Xu, Y. Zhang and Y. Lin To enable fast and effective DNN accelerator development, we propose DNN-Chip Predictor, an analytical performance predictor which can accurately predict DNN accelerators? energy, throughput, and latency prior to their actual implementation.
321 Low-Complexity Fixed-Point Convolutional Neural Networks For Automatic Target Recognition H. Dbouk, H. Geng, C. M. Vineyard and N. R. Shanbhag In order to bring the cost of implementing these networks down, we develop a set of compact network architectures and train them in fixed-point.
322 Accelerating Distributed Deep Learning By Adaptive Gradient Quantization J. Guo et al. In this work, we propose a novel adaptive quantization scheme (AdaQS) to explore the balance between model accuracy and quantization level.
323 Lupulus: A Flexible Hardware Accelerator for Neural Networks A. T. Kristensen, R. Giterman, A. Balatsoukas-Stimming and A. Burg In this work, we present a flexible hardware accelerator for neural networks, called Lupulus, supporting various methods for scheduling and mapping of operations onto the accelerator.
324 Depth Estimation From Single Image Through Multi-Path-Multi-Rate Diverse Feature Extractor W. Lo, C. Chiu and J. Luo This paper proposes a multi-path-multi-rate feature extractor, which can effectively extract multi-scale information to make accurate depth predictions.
325 Object Detection with Color and Depth Images with Multi-Reduced Region Proposal Network and Multi-Pooling J. Lin, C. -. T. Chiu and Y. Cheng We base our network on the Faster R-CNN proposed by Shih et al., and we develop a fast and accurate object detection architecture.
326 Deblurring And Super-Resolution Using Deep Gated Fusion Attention Networks For Face Images C. Yang and L. Chang To solve the problem, we propose a deep gated fusion attention network (DGFAN) to generate a high resolution image without blurring artifacts.
327 Indoor Heading Direction Estimation Using Rf Signals Y. Fan, F. Zhang, C. Wu, B. Wang and K. J. Ray Liu In this paper, we utilize the radio frequency (RF) signals, received from the commercial off-the-shelf (COTS) WiFi devices, to accurately estimate the heading direction in indoor environments.
328 An Improved Selective Active Noise Control Algorithm Based on Empirical Wavelet Transform S. Wen, W. Gan and D. Shi To confront this problem, empirical wavelet transform (EWT) is introduced to simplify the matching and selection of SANC in this paper.
329 Towards Real-Time, Multi-View Video Stereopsis J. Ke, A. J. Watras, J. Kim, H. Liu, H. Jiang and Y. H. Hu We present a real-time, multi-view video stereopsis (RTMVS) algorithm.
330 Ernet Family: Hardware-Oriented Cnn Models For Computational Imaging Using Block-Based Inference C. Huang In this paper, we address these issues by considering the overheads and hardware constraints in advance when constructing CNNs.
331 A DSP Acceleration Framework For Software-Defined Radios On X86 64 G. Georgis, A. Thanos, M. Filo and K. Nikitopoulos This paper presents a DSP acceleration and assessment framework targeting SDR platforms on x86_64 architectures.
332 Fast Single-View 3D Object Reconstruction with Fine Details Through Dilated Downsample and Multi-Path Upsample Deep Neural Network C. Hsu, C. Chiu and C. Kuan To address this issue, we proposed two methods: the dilated downsample block and the multi-path upsample block.
333 Processing Convolutional Neural Networks on Cache J. Vieira, N. Roma, G. Falcao and P. Tom?s In this paper, we propose and assess a novel mechanism that operates at cache level, leveraging both data-proximity and parallel processing capabilities, enabled by dedicated fully-digital vector Functional Units (FUs).
334 Lightweight Hardware Implementation of VVC Transform Block for ASIC Decoder I. Farhat, W. Hamidouche, A. Grill, D. Menard and O. D?forges In this paper, an efficient hardware implementation of all DCT/DST transform types and sizes is proposed.
335 Rgb-D Based Multi-Modal Deep Learning for Face Identification T. Lin, C. Chiu and C. Tang In the proposed architecture, we implement the networks in dual CNN paths for color and depth images separately.
336 A Real Time Implementation of a Bayer Domain Image Deblurring Core for Optical Blur Compensation H. Lee et al. In this letter, we present an implementation of deblurring hardware to mitigate blur incurred by optical aberrations in a real-time manner to increase resolution for mobile camera modules.
337 Self-Attentive Sentimental Sentence Embedding for Sentiment Analysis S. Lin, W. Su, P. Chien, M. Tsai and C. Wang We propose the use of a word-level sentiment bidirectional LSTM in tandem with the self-attention mechanism for sentence-level sentiment prediction.
338 Decidable Variable-Rate Dataflow for Heterogeneous Signal Processing Systems Y. Ma, J. Wu, S. S. Bhattacharyya and J. Boutellier The paper presents the VR-PRUNE model of computation and runtime, and illustrates its applicability to practical signal processing applications by two use cases: an adaptive convolutional neural network, and a predistortion filter for wireless communications.
339 Back-to-Back Butterfly Network, an Adaptive Permutation Network for New Communication Standards H. Harb and C. Chavet In this paper, we introduce an adaptive Back-to-Back Butterfly Network (B2BN) dedicated to next communication standards.
340 1.5GBIT/S 4.9W Hyperspectral Image Encoders on a Low-Power Parallel Heterogeneous Processing Platform O. Ferraz, V. Silva and G. Falcao This work explores the utilization of low-power heterogeneous devices for parallelizing the compute-intensive hyper-spectral and multispectral image compression CCSDS-123 entropy encoders.
341 Design of A Convergence-Aware Based Expectation Propagation Algorithm for Uplink Mimo Scma Systems J. Lin and P. Tsai To further reduce the complexity, a convergence-aware based EPA for uplink MIMO SCMA systems is proposed. Techniques including user termination, antenna termination, and codebook reduction are adopted.
342 Bipartite Belief Propagation Polar Decoding With Bit-Flipping Z. Gong et al. In this paper, we introduce the bit-flipping scheme into the LDPC-like BP (L-BP) decoding and propose two methods to identify the error-prone VNs.
343 Low-Complexity LSTM-Assisted Bit-Flipping Algorithm For Successive Cancellation List Polar Decoder C. Chen, C. Teng and A. A. Wu In this work, we leverage expert knowledge in communication systems and adopt deep learning (DL) techniques to obtain a better solution.
344 Adaptive Normalization for Forecasting Limit Order Book Data Using Convolutional Neural Networks N. Passalis, A. Tefas, J. Kanniainen, M. Gabbouj and A. Iosifidis In this paper we propose a data-driven adaptive normalization layer which is capable of learning the most appropriate normalization scheme that should be applied on the data.
345 Greedy Hybrid Rate Adaptation in Dynamic Wireless Communication Environment Y. Zhao, K. Kang, H. Qian, X. Luo and Y. Jin In this paper, we model the rate selection problem as a multi-armed bandit (MAB) problem and propose an online learning rate adaptation algorithm that learns the channel status from both RSSI and ACK/NACK signals.
346 A WiFi-Based Passive Fall Detection System Y. Hu, F. Zhang, C. Wu, B. Wang and K. J. Ray Liu In this paper, we propose DeFall, a novel WiFi-based environment-independent fall detection system by leveraging the features inherently associated with human falls ? the patterns of speed and acceleration over time.
347 Programmable Dataflow Accelerators: A 5G OFDM Modulation/Demodulation Case Study Y. Wu, P. Wang and J. McAllister This paper describes an FFT accelerator for such a context.
348 Simplified Dynamic SC-Flip Polar Decoding F. Ercan, T. Tonnellier, N. Doan and W. J. Gross In this work, we propose a simple approximation that replaces the transcendental computations of DSCF decoding.
349 Real-Time, Universal, and Robust Adversarial Attacks Against Speaker Recognition Systems Y. Xie, C. Shi, Z. Li, J. Liu, Y. Chen and B. Yuan In this paper, we propose the first real-time, universal, and robust adversarial attack against the state-of-the-art deep neural network (DNN) based speaker recognition system.
350 An Odorant Encoding Machine for Sampling, Reconstruction and Robust Representation of Odorant Identity A. A. Lazar, T. Liu and C. Yeh We solve the sampling problem by developing the Odorant Encoding Machine (OEM), a biomimetic system based on the latest insights in the architectural organization of the fruit fly early olfactory system.
351 FIR Filter Design and Implementation for Phase-Based Processing S. Huang, W. Chen and C. Huang In this paper, we study computation- and memory-efficient finite impulse response filter implementation for CSP.
352 Fixed-Point Optimization of Transformer Neural Network Y. Boo and W. Sung In this study, we quantize the parameters and hidden signals of the Transformer for complexity reduction.
353 A Fifo Based Accelerator for Convolutional Neural Networks V. Panchbhaiyye and T. Ogunfunmi In this paper we present an architecture which takes a novel approach to compute convolution results using row-wise inputs as opposed to traditional tile-based processing.
354 Soft-Output Finite Alphabet Equalization for mmWave Massive MIMO O. Casta?eda, S. Jacobsson, G. Durisi, T. Goldstein and C. Studer In this work, we improve upon finite-alphabet equalization by performing unbiased estimation and soft-output computation for coded systems.
355 Diversity and Sparsity: A New Perspective on Index Tracking Y. Zheng, T. M. Hospedales and Y. Yang We introduce the first index tracking method that explicitly optimises both diversity and sparsity in a single joint framework.
356 Sparse Beamspace Equalization for Massive MU-MIMO MMWave Systems S. H. Mirfarshbafan and C. Studer We propose equalization-based data detection algorithms for all-digital millimeter-wave (mmWave) massive multiuser multiple-input multiple-out (MU-MIMO) systems that exploit sparsity in the beamspace domain to reduce complexity.
357 Threshold-Adjusted ORB Strategies with Genetic Algorithm and Protective Closing Strategy on Taiwan Futures Market J. Syu, M. Wu, C. Chen and J. Ho In this paper, we adjust thresholds through historical data to enhance profitability, and design protective closing strategy to prevent unacceptable losses.
358 Can every analog system be simulated on a digital computer? H. Boche and V. Pohl This paper shows that there exist very simple linear time-invariant (LTI) systems which can not be simulated on a Turing machine.
359 Low-Complexity Compressed Alignment-Aided Compressive Analysis for Real-Time Electrocardiography Telemonitoring Y. Pua, C. Chou and A. A. Wu In this paper, we propose a new compressed alignment-aided compressive analysis (CA-CA) framework that enables simple alignment and low-complexity requirements.
360 Reduced-Complexity Singular Value Decomposition For Tucker Decomposition: Algorithm And Hardware X. Hu, C. Deng and B. Yuan In this paper, we propose a reduced-complexity SVD (Singular Vector Decomposition) scheme, which serves as the key operation in Tucker decomposition.
361 An Early Termination Scheme for Successive Cancellation List Decoding of Polar Codes H. Lee, Y. Pao, C. Chi, H. Lee and Y. Ueng In order to minimize the decoding period and the response time for Polar Codes, an early termination (ET) scheme based on additional check points (ACPs) is proposed in this work.
362 Low-Latency Lightweight Streaming Speech Recognition with 8-Bit Quantized Simple Gated Convolutional Neural Networks J. Park, X. Qian, Y. Jo and W. Sung In this paper, we propose a low-latency on-device speech recognition system with a simple gated convolutional network (SGCN).
363 Shape From Bandwidth: Central Projection Case G. Elhami, A. J. Scholefield and M. Vetterli We model the problem as a central projection operation in 2D and propose a two step approach for finding the surface from the observed image.
364 Sequential Deep Unrolling With Flow Priors For Robust Video Deraining X. Xue, Y. Ding, P. Mu, L. Ma, R. Liu and X. Fan A sequential deep unrolling framework is substantially presented by solving this model based on optimization techniques.
365 A Fast and Accurate Super-Resolution Network Using Progressive Residual Learning H. Liu, Z. Lu, W. Shi and J. Tu In this work, a lightweight network using progressive residual learning for SISR (PRLSR) is proposed to address this issue.
366 REV-AE: A Learned Frame Set for Image Reconstruction S. Li, Z. Zheng, W. Dai, J. Zou and H. Xiong In this paper, we propose a reversible autoencoder (Rev-AE) with this extended non-linear lifting scheme to improve image reconstruction.
367 Decomposed Cyclegan for Single Image Deraining With Unpaired Data K. Han and X. Xiang Inspired by adopting unpaired data in task of translation, in this paper we present a new method for rain removal using unpaired data.
368 Slicenet: Slice-Wise 3D Shapes Reconstruction from Single Image Y. Wu, Z. Sun, Y. Song, Y. Sun and J. Shi In this paper, we propose SliceNet, sequentially generating 2D slices of 3D shapes with shared 2D deconvolution parameters.
369 Sight to Sound: An End-to-End Approach for Visual Piano Transcription A. S. Koepke, O. Wiles, Y. Moses and A. Zisserman In this work, we address the problem of transcribing piano music from visual data alone.
370 Exocentric to Egocentric Image Generation Via Parallel Generative Adversarial Network G. Liu, H. Tang, H. Latapie and Y. Yan In this paper, we investigate exocentric (third-person) view to egocentric (first-person) view image generation.
371 Focus on Semantic Consistency for Cross-Domain Crowd Understanding T. Han, J. Gao, Y. Yuan and Q. Wang However, we found that a mass of estimation errors in the background areas impede the performance of the existing methods. In this paper, we propose a domain adaptation method to eliminate it.
372 Improved Real-Time Visual Tracking via Adversarial Learning H. Zhong, X. Yan, Y. Jiang and S. Xia In this paper, we attempt to combine the advantages from both methods and propose an improved real-time visual tracking algorithm via adversarial learning to get a more balanced result in accuracy and tracking speed.
373 Spatial-Temporal Feature Aggregation Network For Video Object Detection Z. Chen, W. Li, C. Fei, B. Liu and N. Yu In this paper, we propose a novel spatial-temporal feature aggregation network to deal with this issue.
374 Using Panoramic Videos for Multi-Person Localization and Tracking In A 3D Panoramic Coordinate F. Yang, F. Li, Y. Wu, S. Sakti and S. Nakamura In this work, we propose an effective and efficient approach at a low cost.
375 Position Constraint Loss For Fashion Landmark Estimation H. Liu, M. Song, W. Shi and X. Li To alleviate these issues, we propose Position Constraint Loss (PCLoss) to constrain error landmark locations by utilizing the position relationship of landmarks.
376 Receptive Field Pyramid Network for Object Detection F. Wu, A. J. Ma, Y. Pan, Y. Gao and X. Yan To overcome this limitation and carry out better object detection, we design a novel network named Receptive Field Pyramid Network (RFPN).
377 Robust Visual Tracking with Context-Based Active Occlusion Recognition Y. Gu, Y. Qiao, K. Xu, H. Xu and X. Fang In this paper, we propose a context-based active occlusion recognition framework that can be integrated with various tracking approaches.
378 Leveraging Ordinal Regression With Soft Labels For 3d Head Pose Estimation From Point Sets S. Xiao, N. Sang, X. Wang and X. Ma In contrast to existing approaches that take 2D depth image as input, we propose a novel deep regression architecture called Head PointNet, which consumes 3D point sets derived from a depth image describing the visible surface of a head.
379 Solving Missing-Annotation Object Detection with Background Recalibration Loss H. Zhang, F. Chen, Z. Shen, Q. Hao, C. Zhu and M. Savvides In this paper, we introduce a superior solution called Background Recalibration Loss (BRL) that can automatically re-calibrate the loss signals according to the pre-defined IoU threshold and input image.
380 Face Feature Recovery via Temporal Fusion for Person Search C. Fan, C. Liu, K. Wang, J. Jhan, Y. F. Wang and J. Chen To address the issue, we propose a unique framework of “Face Feature Recovery via Temporal Fusion” to synthesize virtual facial features by observing both temporal and contextual information.
381 Edgefool: an Adversarial Image Enhancement Filter A. S. Shamsabadi, C. Oh and A. Cavallaro In this paper, we propose EdgeFool, an adversarial image enhancement filter that learns structure-aware adversarial perturbations.
382 Facial Feature Embedded Cyclegan For Vis-Nir Translation H. Wang, H. Zhang, L. Yu, L. Wang and X. Yang Inspired by the CycleGAN, this paper presents a method aiming to translate between VIS and NIR face images.
383 Deep Image Deblurring Using Local Correlation Block W. Su, Y. Yuan and Q. Wang In this paper, we propose a local correlation block (LCBlock), which can adjust the weights of features adaptively according to the blurry inputs.
384 Global Structure Graph Guided Fine-Grained Vehicle Recognition C. Wang, H. Fu and H. Ma In this paper, we propose an approach that introduces the structure graph into consideration to learn distinguishing representations for vehicle recognition.
385 Triplet Loss Feature Aggregation for Scalable Hash W. Jia, L. Li, Z. Li, S. Zhao and S. Liu In this paper, we propose a novel content-based video segmentation identification scheme that is invariant to the underlying codec and operational bit rates.
386 HDMFH: Hypergraph Based Discrete Matrix Factorization Hashing for Multimodal Retrieval J. Gao, W. Zhang, Z. Chen and F. Zhong To address these issues, in this paper, we propose a novel cross-modal hashing method, named Hypergraph Based Discrete Matrix Factorization Hashing (HDMFH), for multimodal retrieval.
387 Multi-Scale Deep Feature Fusion for Vehicle Re-Identification Y. Cheng, C. Zhang, K. Gu, L. Qi, Z. Gan and W. Zhang In this paper, we propose a novel Multi-Scale Deep Feature Fusion Network (MSDeep) to conduct both multi-scale and multi-level features for precise vehicle re-id.
388 Crowdsourcing-Based Ranking Aggregation for Person Re-Identification Y. Yu, C. Liang, W. Ruan and L. Jiang To this end, this paper proposes a crowdsourcing-based ranking aggregation to adaptively fuse multiple ranking lists for re-ID problem.
389 Deep Multi-Region Hashing Q. Zhou, X. Nie, Y. Shi, X. Liu and Y. Yin To address this issue, in this paper, we proposed a novel Deep Multi-Region Hashing (DMRH) method to fully utilize the semantic details, which uses overlapping N ? N regions of an image to learn N2 hash codes for getting a final hash code.
390 Semantic Augmentation Hashing for Zero-Shot Image Retrieval F. Zhong, Z. Chen, G. Min and F. Xia In this paper, we propose a novel Semantic Augmentation Hashing (SAH) for zero-shot image retrieval.
391 End-To-End Generation of Talking Faces from Noisy Speech S. E. Eskimez, R. K. Maddox, C. Xu and Z. Duan In this work, we propose an end-to-end (no pre- or post-processing) system that can generate talking faces from arbitrarily long noisy speech.
392 Unsupervised Image-to-Image Translation Via Fair Representation of Gender Bias S. Hwang and H. Byun In this work, we propose a framework of unsupervised Image-to-Image translation that learns a fair representation by separating the latent space of our model into two purposes: 1) Target Attribute Editing, 2) Gender Preserving.
393 Video Frame Interpolation Via Exceptional Motion-Aware Synthesis M. Park, S. Lee and Y. M. Ro In this paper, we propose a novel video frame interpolation method via exceptional motion-aware synthesis, in which accurate optical flow could be estimated even with exceptional motion patterns.
394 Look Globally, Age Locally: Face Aging With an Attention Mechanism H. Zhu, Z. Huang, H. Shan and J. Zhang To address this deficiency, this paper introduces an Attention Conditional GANs (AcGANs) approach for face aging, which utilizes attention mechanism to only alert the regions relevant to face aging.
395 Design-Gan: Cross-Category Fashion Translation Driven By Landmark Attention Y. Lang, Y. He, J. Dong, F. Yang and H. Xue In this paper, we propose a novel approach, called DesignGAN, that utilizes the landmark guided attention and a similarity constraint mechanism to achieve fashion cross-category translation.
396 Intensity-Image Reconstruction for Event Cameras Using Convolutional Neural Network Y. Chen, W. Chen, X. Cao and Q. Hua In this paper, “event frames” are recovered from event streams in an attenuation method and they are fed into the U-net network to generate intensity images.
397 Colour Compression of Plenoptic Point Clouds Using Raht-Klt with Prior Colour Clustering and Specular/Diffuse Component Separation M. Krivokuca and C. Guillemot In the current paper, we demonstrate that the best-proposed RAHT extension, RAHT-KLT, can be improved by performing a prior subdivision of the plenoptic point cloud into clusters based on similar colour values, followed by a separation of each cluster into specular and diffuse components, and coding each component separately with RAHT-KLT.
398 Super-Resolution of 3D Color Point Clouds Via Fast Graph Total Variation C. Dinesh, G. Cheung and I. V. Bajic In this paper, we propose a fast super-resolution (SR) algorithm for color 3D point clouds.
399 Deep Monocular Video Depth Estimation Using Temporal Attention H. Ren, M. El-khamy and J. Lee In this paper, we propose an end-to-end monocular video depth estimation network based on temporal attention.
400 Robust Full-Fov Depth Estimation in Tele-Wide Camera System K. Guo, S. Song, S. Chang, T. Kim, S. Han and I. Kim In this paper, to address the above problems we propose a hierarchical hourglass network for robust full-FoV depth estimation in tele-wide camera system, which combines the robustness of traditional stereo-matching methods with the accuracy of DNN.
401 Manet: Multi-Scale Aggregated Network For Light Field Depth Estimation Y. Li, L. Zhang, Q. Wang and G. Lafruit We present a novel end-to-end network, MANet, for light field depth estimation.
402 EPI-Neighborhood Distribution Based Light Field Depth Estimation J. Li and X. Jin In this paper, a novel depth estimation algorithm tackling foreground occlusion is proposed based on the neighborhood distribution in the sheared epipolar images (EPIs).
403 Stochastic Multi-Scale Aggregation Network for Crowd Counting M. Wang, H. Cai, J. Zhou and M. Gong This paper presents a novel end-to-end stochastic multi-scale aggregation network (SMANet) which carefully addresses these issues.
404 MDR-SURV: A Multi-Scale Deep Learning-Based Radiomics for Survival Prediction in Pulmonary Malignancies P. Afshar, A. Oikonomou, K. N. Plataniotis and A. Mohammadi In this work, we propose a Multi-scale Deep learning-based Radiomics model, referred to as “MDR-SURV” that exploits the information from positron emission tomography/computed tomography (PET/CT) images, combined with other clinical factors, to predict the overall survival (OS).
405 Learning a Generic Adaptive Wavelet Shrinkage Function for Denoising T. Alt and J. Weickert To reduce this gap, we introduce a generic wavelet shrinkage function for denoising which is adaptive to both the wavelet scales as well as the noise standard deviation.
406 Multi-Scale Residual Network for Image Classification X. Zhong, O. Gong, W. Huang, J. Yuan, B. Ma and R. W. Liu In this paper, we propose the Multi-Scale Residual (MSR) module that integrates multi-scale feature maps of the underlying information to the last layer of Convolutional Neural Network.
407 Deep Multi-Scale Gabor Wavelet Network for Image Restoration H. Dong, X. Zhang, Y. Guo and F. Wang In this paper, we propose a Multiscale Gabor Wavelet Network (MsGWN) for image restoration.
408 Residual Attention Network for Wavelet Domain Super-Resolution J. Liu, Y. Xie, H. Song, W. Yuan and L. Ma In this paper, we propose a novel network with better textural details in wavelet domain, which is composed of a feature extract layer, residual channel attention groups (RCAG) and a residual up-sampling layer based on inverse discrete wavelet transform.
409 An Adaptive Linear Estimator Based Approach to Bi-Directional Motion Compensated Prediction B. Li, J. Han and K. Rose This paper proposes a novel bi-directional motion compensation mode that efficiently utilizes the motion information that is already available to the decoder, without recourse to extensive search.
410 Spherical Video Coding with Geometry and Region Adaptive Transform Domain Temporal Prediction B. Vishwanath and K. Rose To account for such variations, we propose geometry and region adaptive TDTP that is tailored to spherical videos.
411 Versatile Video Coding and Super-Resolution for Efficient Delivery of 8k Video with 4k Backward-Compatibility C. Bonnineau, W. Hamidouche, J. Travers and O. D?forges In this paper, we propose, through an objective study, to compare and evaluate the performance of different coding approaches allowing the delivery of an 8K video signal with 4K backward-compatibility on broadcast networks.
412 Alternative Half-Sample Interpolation Filters for Versatile Video Coding A. Henkel et al. In this paper, alternative half-luma-sample interpolation filters are proposed.
413 Just Noticeable Distortion Based Perceptually Lossless Intra Coding X. Shen, X. Zhang, S. Wang, S. Kwong and G. Zhu In this paper, a just noticeable distortion (JND) guided perceptually lossless coding framework is proposed for Versatile Video Coding (VVC) intra coding.
414 Efficient Deep Learning-Based Lossy Image Compression Via Asymmetric Autoencoder and Pruning J. Kim, J. Choi, J. Chang and J. Lee In this paper, we propose efficient lossy image compression methods based on asymmetric autoencoder and decoder pruning.
415 Deriving Compact Feature Representations Via Annealed Contraction M. A. Shah and B. Raj To this end we propose a technique that shrinks a layer by an iterative process in which neurons are removed from the and network is fine tuned.
416 Fast Clustering With Co-Clustering Via Discrete Non-Negative Matrix Factorization for Image Identification F. Nie, S. Pei, R. Wang and X. Li To address this problem, a novel clustering method called fast clustering with co-clustering via discrete non-negative matrix factorization, is proposed.
417 Compressive Adaptive Bilateral Filtering P. Nair, R. G. Gavaskar and K. N. Chaudhury We propose a fast algorithm for an adaptive variant of the classical bilateral filter, where the range kernel is allowed to vary from pixel to pixel.
418 Attention Mechanism Enhanced Kernel Prediction Networks for Denoising of Burst Images B. Zhang, S. Jin, Y. Xia, Y. Huang and Z. Xiong In this paper, attention mechanism enhanced kernel prediction networks (AME-KPNs) are proposed for burst image denoising, in which, nearly cost-free attention modules are adopted to first refine the feature maps and to further make a full use of the inter-frame and intra-frame redundancies within the whole image burst.
419 Image Restoration Via Data-Dependent Proximal Averaged Optimization P. Mu, J. Chen, R. Liu, W. Zhong, X. Fan and Z. Luo To partially address the above issues, we develop a Data-dependent Proximal Averaged (DPA) paradigm through optimizing objective and data-dependent feasibility constraint for the challenging Image Restoration (IR) tasks.
420 Multi-Way Multi-View Deep Autoencoder for Image Feature Learning with Multi-Level Graph Regularization Z. Fang, S. Zhou, X. Li and H. Zhu In this paper, we propose a multi-way deep autoencoder for multi-view feature learning to explore the deep consensus structure and reconcile the efficiency of encoding process meanwhile.
421 Exposure Interpolation Via Hybrid Learning C. Zheng, Z. Li, Y. Yang and S. Wu A natural question would be “Is there any space for conventional methods on these problems?” In this paper, exposure interpolation is taken as an example to answer this question and the answer is “Yes”.
422 Learning Spatio-Temporal Representations With Temporal Squeeze Pooling G. Huang and A. G. Bors In this paper, we propose a new video representation learning method, named Temporal Squeeze (TS) pooling, which can extract the essential movement information from a long sequence of video frames and map it into a set of few im?ages, named Squeezed Images.
423 Fine-Grained Giant Panda Identification R. Ding, L. Wang, Q. Zhang, Z. Niu, N. Zheng and G. Hud To address these challenges, we propose the Feature-Fusion Convolutional Neural Network with Patch Detector (FFCNN-PD) algorithm, which exploits the discriminative local patches and builds a hierarchical representation generated by fusing both global and local features.
424 Learning from Dances: Pose-Invariant Re-Identification for Multi-Person Tracking H. Ho, M. Shim and D. Wee With the goal of learning pose-invariant representations, we propose an end-to-end deep learning framework Sparse-Temporal ReID Network.
425 Learning Fractional Orthogonal Latent Consistent Features for Face Hallucination and Recognition Y. Yuan, J. Li, Y. Li, J. Qiang and B. Li In this paper, we propose a novel FH method via fractional orthogonal latent consistent features that we call fractional orthogonal partial least squares based FH (FOPLSFH).
426 Sparse Modeling on Distributed Encryption Data Y. BANDOH, T. NAKACHI and H. KIYA In this paper, we construct an analysis model for data encrypted with the random unitary transform by deriving a LASSO solution for encrypted data.
427 S-DOD-CNN: Doubly Injecting Spatially-Preserved Object Information for Event Recognition H. Lee, S. Eum and H. Kwon We present a novel event recognition approach called Spatially-preserved Doubly-injected Object Detection CNN (S-DOD-CNN), which incorporates the spatially preserved object detection information in both a direct and an indirect way.
428 Angular Discriminative Deep Feature Learning for Face Verification B. Wu and H. Wu In order to bridge the gap between training and testing, we require that the intra-class cosine similarity of the inner-product layer before softmax loss is larger than a margin in the training step, accompanied by the supervision signal of softmax loss.
429 3d Deformation Signature for Dynamic Face Recognition A. E. Rahman Shabayek, D. Aouada, K. Cherenkova, G. Gusev and B. Ottersten This work proposes a novel 3D Deformation Signature (3DS) to represent a 3D deformation signal for 3D Dynamic Face Recognition.
430 ASR is All You Need: Cross-Modal Distillation for Lip Reading T. Afouras, J. S. Chung and A. Zisserman The goal of this work is to train strong models for visual speech recognition without requiring human annotated ground truth data.
431 Color Stabilization for Multi-Camera Light-Field Imaging O. V. Thanh, T. Canham, J. Vazquez-Corral, R. G. Rodr?guez and M. Bertalm?o In this work we adapt and extend to the light-field scenario a color stabilization method previously proposed for standard multi-camera shoots, and demonstrate experimentally that it provides an improvement over the state-of-the-art techniques for light-field imaging.
432 Learning Spatio-Temporal Convolutional Network for Real-Time Object Tracking H. Chen, X. Xing and X. Xu In this paper we focus on making using of the rich information in latest consecutive frames to improve the feature representation of initial template frame.
433 Learned Lossless Image Compression with A Hyperprior and Discretized Gaussian Mixture Likelihoods Z. Cheng, H. Sun, M. Takeuchi and J. Katto This paper generalizes the hyperprior from lossy model to lossless compression, and proposes a L2-norm term into the loss function to speed up training procedure.
434 Variable Bitrate Image Compression with Quality Scaling Factors T. Chen and Z. Ma In this paper, we propose to embed a set of quality scaling factors (SFs) into learned image compression network, by which we can encode images across an entire bitrate range with a single model.
435 Binary Probability Model for Learning Based Image Compression T. LADUNE, P. PHILIPPE, W. HAMIDOUCHE, L. ZHANG and O. D?FORGES In this paper, we propose to enhance learned image compression systems with a richer probability model for the latent variables.
436 Improved Probability Modelling for Exception Handling in Lossless Screen Content Coding T. Strutz This paper proposes an enhanced version of this exception handling coding.
437 Spatially Adaptive Intra Mode Pre-Selection for ERP 360 Video Coding I. Storch, B. Zatt, L. Agostini, G. Correa, L. A. da Silva Cruz and D. Palomino In this work, we propose a spatially adaptive HEVC intra mode pre-selection for equirectangular (ERP) 360 video coding.
438 Semi-Regular Geometric Kernel Encoding & Reconstruction for Video Compression X. Jiang, C. Yang, G. Cheung and S. Takamura In this paper, we extract a best-fitting “semi-regular” geometric structure from a target spatial region in a frame group, which is encoded separately as a unified signal predictor for these frames.
439 Leveraging Cuboids for Better Motion Modeling in High Efficiency Video Coding A. Ahmmed, M. Murshed and M. Paul Leveraging on the advantages of cuboids, in this paper, we propose to discover homogeneous motion regions and their associated motion based on cuboids.
440 Adversarial Video Compression Guided by Soft Edge Detection S. Kim et al. We propose a video compression framework using conditional Generative Adversarial Networks (GANs).
441 Compressing Flow Fields with Edge-Aware Homogeneous Diffusion Inpainting F. Jost, P. Peter and J. Weickert In spite of the fact that efficient compression methods for dense two-dimensional flow fields would be very useful for modern video codecs, hardly any research has been performed in this area so far. Our paper addresses this problem by proposing the first lossy diffusion-based codec for this purpose.
442 Adaptive Resolution Change Using Uncoded Areas and Dictionary Learning-Based Super-Resolution in Versatile Video Coding J. Schneider, J. Sauer and C. Rohlfing This contribution introduces an ARC concept using un-coded areas within a frame of a video sequence and a Dictionary Learning (DL)-based Super-Resolution (SR) scheme.
443 RDE-MOGA: Automatic Selection of Rate-Distortion-Energy Control Points for Video Encoders Using Muti-Objetive Genetic Algorithm ?. D. Machado, M. S. de Aguiar, M. Porto, G. Corr?a, D. Palomino and B. Zatt In this work we propose the RDE-MOGA, an multi-objective genetic algorithm capable of finding energetically efficient configurations for the HEVC encoder and replacing the current sensitivity analysis methodologies in the development of energy controllers.
444 A Connected Auto-Encoders Based Approach for Image Separation with Side Information: With Applications to Art Investigation W. Pu, B. Sober, N. Daly, C. Higgitt, I. Daubechies and M. R. D. Rodrigues In this paper, we propose a new architecture based on the use of ?connected? auto-encoders in order to separate mixed X-ray images acquired from double-sided paintings, where in addition to the mixed X-ray image one can also exploit the two RGB images associated with the front and back of the painting.
445 Self-Supervised Adversarial Training K. Chen et al. With these views, we introduce self-supervised learning to against adversarial examples in this paper.
446 Gray-Scale Image Colorization Using Cycle-Consistent Generative Adversarial Networks with Residual Structure Enhancer M. M. Johari and H. Behroozi Since one can consider the gray-scale and colorful images as two separate domains, we propose a two-stage cycle-consistent network architecture to produce convincible images.
447 All You Need is a Second Look: Towards Tighter Arbitrary Shape Text Detection M. Cao and Y. Zou To address these problems, we innovatively propose a two-stage segmentation based arbitrary text detector named NASK (Need A Second looK).
448 Compare Learning: Bi-Attention Network for Few-Shot Learning L. Ke, M. Pan, W. Wen and D. Li In this paper, we propose a novel approach named Bi-attention network to compare the instances, which can measure the similarity between embeddings of instances precisely, globally and efficiently.
449 Arnet: Attention-Based Refinement Network for Few-Shot Semantic Segmentation R. Li, H. Liu, Y. Zhu and Z. Bai In this paper, we propose an Attention-based Refinement Network (ARNet) for few-shot semantic segmentation, which consists of three branches: the guidance branch, the segmentation branch and the refinement branch.
450 Lightdet: A Lightweight and Accurate Object Detection Network Q. Tang, J. Li, Z. Shi and Y. Hu In this paper, we present a lightweight object detector, named LightDet, to address this dilemma.
451 Self-Supervised Deep Learning for Fisheye Image Rectification C. Chao, P. Hsu, H. Lee and Y. F. Wang To rectify fisheye distortion from a single image, we advance self-supervised learning strategies and propose a unique deep learning model of Fisheye GAN (FE-GAN).
452 Sketchppnet: A Joint Pixel and Point Convolutional Neural Network For Low Resolution Sketch Image Recognition X. Zhu, Y. Xiao, Y. Zheng, G. Tan and S. Zhou To solve this problem, we propose a joint pixel and point convolutional neural network for LR sketch image recognition.
453 All In One Network for Driver Attention Monitoring D. Yang et al. To address this problem, we propose a multi-task learning CNN framework which simultaneously solve these tasks.
454 Unsupervised Domain Adaptation for Semantic Segmentation with Symmetric Adaptation Consistency Z. Li, R. Togo, T. Ogawa and M. Haseyama In this paper, we utilize adversarial learning and semi-supervised learning simultaneously to solve the task of unsupervised domain adaptation in semantic segmentation.
455 IQ-STAN: Image Quality Guided Spatio-Temporal Attention Network for License Plate Recognition C. Zhang, Q. Wang and X. Li In order to tackle this issue, a novel deep multi-task learning-based method is proposed in this paper by introducing contextual information in multiple license plate frames.
456 Weakly Supervised Semantic Segmentation For Remote Sensing Hyperspectral Imaging E. Moliner, L. S. Romero and V. Vilaplana This paper studies the problem of training a semantic segmentation neural network with weak annotations, in order to be applied in aerial vegetation images from Teide National Park.
457 Social Data Assisted Multi-Modal Video Analysis For Saliency Detection J. Xia et al. In this paper, we propose a novel learning-based multi-modal method for optimizing user-oriented video analysis.
458 View-Angle Invariant Object Monitoring Without Image Registration X. Zhang, C. Huo and C. Pan To address the above difficulties, a novel object-specific change detection approach is proposed for object monitoring in this paper.
459 Hierarchical Sequence Representation with Graph Network D. Chen, X. Wu, J. Dong, Y. He, H. Xue and F. Mao In this paper, we propose a novel video classification method based on a deep convolutional graph neural network (DCGN).
460 Multi Image Depth from Defocus Network with Boundary Cue for Dual Aperture Camera G. Song, Y. Kim, K. Chun and K. M. Lee In this paper, we estimate depth information using two defocused images from dual aperture camera.
461 Height and Weight Estimation from Unconstrained Images C. Y. Altinigne, D. Thanou and R. Achanta We present a deep learning scheme that relies on simultaneous prediction of human silhouettes and skeletal joints as strong regularizers that improve the prediction of attributes such as height and weight.
462 Sampling Strategies for GAN Synthetic Data B. Bhattarai, S. Baek, R. Bodur and T. Kim To this end, we propose to maximally utilise the parameters learned during training of the GAN itself.
463 Auglabel: Exploiting Word Representations to Augment Labels for Face Attribute Classification B. Bhattarai, R. Bodur and T. Kim In this paper, we present a simple, yet effective novel method to generate fixed dimensional labels with continuous values for images by exploiting the word2vec ? semantic representations ? of the existing categorical labels.
464 Multi-Task Center-Of-Pressure Metrics Estimation from Skeleton Using Graph Convolutional Network C. Du, S. Graham, S. Jin, C. Depp and T. Nguyen In this paper, we propose an end-to-end framework to estimate the COP path length and the COP positions from the 3D skeleton, utilizing the spatial-temporal features learned by graph convolutional networks.
465 Regression Before Classification for Temporal Action Detection C. Jin, T. Zhang, W. Kong, T. Li and G. Li In this paper, we propose to eliminate this inconsistency by making two modifi-cations to the action classifier: 1) redirecting the classification loss to the refined proposal, and 2) rearranging the location regressor before the action classifier so that the feature of the refined proposal is fed to the classifier.
466 Multi-Task Learning in Autonomous Driving Scenarios Via Adaptive Feature Refinement Networks M. Zhai, X. Xiang, N. Lv and A. E. Saddik In this paper, we combine an adaptive feature refinement module and a unified framework for joint learning of optical flow, depth and camera pose estimation in an unsupervised manner.
467 A Real-Time Deep Network for Crowd Counting X. Shi, X. Li, C. Wu, S. Kong, J. Yang and L. He In this paper, we propose a compact convolutional neural network for crowd counting which learns a more efficient model with a small number of parameters.
468 Drift Detection and Correction Post-Tracking T. Ghoniemy and M. A. Amer This paper proposes a method that first detects, at each frame, if a tracker tends to drift by analyzing saliency features of the output BB of a tracker, and then applies automatic seeded object segmentation on the BB to correct the drift once detected.
469 Interpretable Self-Attention Temporal Reasoning for Driving Behavior Understanding Y. Liu, Y. Hsieh, M. Chen, C. -. H. Yang, J. Tegner and Y. -. J. Tsai We proposed a perturbation-based visual explanation method to inspect the models? performance visually.
470 Non-Uniform Video Time-Lapse Method Based on Motion Scenario and Stabilization Constraint K. Guo et al. More specific, we introduce an advanced Markov chain (MC) model, in which smoothed camera trajectory, FOV loss constraint of VS, camera motion scenario, and sampling interval similarity between consecutive frames are encoded as potential functions.
471 Key Action and Joint CTC-Attention based Sign Language Recognition H. Li, L. Gao, R. Han, L. Wan and W. Feng In this paper, we propose to hierarchically search key actions by a pyramid BiLSTM.
472 Learning Geometric Features with Dual?stream CNN for 3D Action Recognition T. Huynh-The, C. Hua, N. A. Tu and D. Kim This paper introduces a deep network configured by two parallel streams of convolutional stacks for fully learning the deep intra-frame joint associations and inter-frame joint correlations, wherein the structure of each stream is learned from Inception-v3.
473 A Deep Learning Approach to Object Affordance Segmentation S. Thermos, P. Daras and G. Potamianos In this paper, we propose a novel approach that exploits the spatio-temporal nature of human-object interaction for affordance segmentation.
474 Multi-View Shape Estimation of Transparent Containers A. Xompero, R. Sanchez-Matilla, A. Modas, P. Frossard and A. Cavallaro In this paper, we propose a method for jointly localising container-like objects and estimating their dimensions using two wide-baseline, calibrated RGB cameras.
475 Rethinking Temporal-Related Sample for Human Action Recognition J. Wang, S. Li, Z. Duan and Z. Yuan In this paper, our motivation is to address this issue by utilizing temporal information more effectively.
476 FDDWNet: A Lightweight Convolutional Neural Network for Real-Time Semantic Segmentation J. Liu, Q. Zhou, Y. Qiang, B. Kang, X. Wu and B. Zheng This paper introduces a lightweight convolutional neural network, called FDDWNet, for real-time accurate semantic segmentation.
477 Complex Pairwise Activity Analysis Via Instance Level Evolution Reasoning S. Paul, C. Torres, S. Chandrasekaran and A. K. Roy-Chowdhury This work introduces a novel method that jointly exploits relational information between pairs of objects and temporal dynamics of each object.
478 Scene Text Recognition with Temporal Convolutional Encoder X. Du, T. Ma, Y. Zheng, H. Ye, X. Wu and L. He In this paper, we study text recognition framework by considering the long-term temporal dependencies in the encoder stage.
479 Enhanced Action Tubelet Detector for Spatio-Temporal Video Action Detection Y. Wu, H. Wang, S. Wang and Q. Li To this end, a single stream network named enhanced action tubelet (EAT) detector is proposed in this work based on RGB stream.
480 Secure Face Recognition in Edge and Cloud Networks: From the Ensemble Learning Perspective Y. Wang and T. Nakachi In this paper, we develop a secure face recognition framework to orchestrate sparse coding in edge and cloud networks.
481 Low Complexity Single Image Super-Resolution with Channel Splitting and Fusion Network M. Zou, J. Tang and G. Wu In this paper, we propose a low complexity solution based on channel splitting and fusion network (CSFN) to address this problem.
482 Learning Spectral-Spatial Prior Via 3DDNCNN for Hyperspectral Image Deconvolution X. Wang, J. Chen, C. Richard and D. Brie In this paper, we use the alternating direction method of multipliers (ADMM) to decompose the optimization problem into iterative subproblems where the prior only appears in a denoising subproblem.
483 Dynamically Modulated Deep Metric Learning for Visual Search D. Manandhar, M. Bastan and K. Yap This paper proposes dynamically modulated metric learning (DMML) for learning a tiered similarity space to perform visual search.
484 Deep Residual Network for MSFA Raw Image Denoising Z. Pan, B. Li, H. Cheng and Y. Bao To overcome these challenges, we propose a new deep residual network designed to account for the uniqueness of MSFA mosaic patterns.
485 MSPNET: Multi-Supervised Parallel Network for Crowd Counting B. Wei, Y. Yuan and Q. Wang In this paper, we propose a multi-supervised parallel network (MSPNet) to achieve high accuracy of crowd counting and generate high-quality density maps.
486 Video Question Generation via Semantic Rich Cross-Modal Self-Attention Networks Learning Y. Wang, H. Su, C. Chang, Z. Liu and W. H. Hsu We introduce a novel task, Video Question Generation (Video QG).
487 Blind Hyperspectral Unmixing using Dual Branch Deep Autoencoder with Orthogonal Sparse Prior Z. Dou, K. Gao, X. Zhang, H. Wang and J. Wang In this paper, we propose a dual branch autoencoder with a novel sparse prior to simultaneously extract endmembers and abundances from the raw HSI.
488 Classification of Depth and Surface Edges with Deep Features Z. Li and X. Wu In this paper we study the problem of automatic classification of the two types of edges.
489 Learning to Characterize Adversarial Subspaces X. Mao, Y. Chen, Y. Li, Y. He and H. Xue To solve this problem, we propose a novel adversarial detection method which identifies adversaries by adaptively learning reasonable metrics to characterize adversarial subspaces.
490 Video Deblurring Via 3d CNN and Fourier Accumulation Learning F. Yang, L. Xiao and J. Yang In this paper, we propose a simple yet effective Fourier accumulation embedded 3D convolutional encoder-decoder network for video deblurring.
491 Enhanced Non-Local Cascading Network with Attention Mechanism for Hyperspectral Image Denoising H. Ma, G. Liu and Y. Yuan In this paper, a novel HSIs denoising algorithm based on an enhanced non-local cascading network with attention mechanism (ENCAM) is proposed, which can extract the joint spatial-spectral feature more effectively.
492 Quantized Tensor Robust Principal Component Analysis A. Aidini, G. Tsagkatakis and P. Tsakalides In this paper, we introduce a tensor robust principal component analysis algorithm in order to recover a tensor with real-valued entries from a partly observed set of quantized and sparsely corrupted entries.
493 A New Perspective for Flexible Feature Gathering in Scene Text Recognition Via Character Anchor Pooling S. Long, Y. Guan, K. Bian and C. Yao To tackle these issues, we propose a pair of coupling modules, termed as Character Anchoring Module (CAM) and Anchor Pooling Module (APM), to extract high-level semantics from two-dimensional space to form feature sequences.
494 Hybrid Active Contour Driven by Double-Weighted Signed Pressure Force for Image Segmentation X. Fu, B. Fang, M. Zhou and J. Li In this paper, we proposed a novel hybrid active contour driven by double-weighted signed pressure force method for image segmentation.
495 Neural Coding Strategies for Event-Based Vision Data S. Harrigan, S. Coleman, D. Kerr, P. Yogarajah, Z. Fang and C. Wu This paper introduces three different neural coding scheme formations for event-based vision data which are designed to emulate the neural behaviour exhibited by neurons under stimuli.
496 Camera Configuration Design in Cooperative Active Visual 3d Reconstruction: A Statistical Approach Q. An and Y. Shen In this paper, we propose a statistical framework for the active visual 3D reconstruction.
497 Hand-3d-Studio: A New Multi-View System for 3d Hand Reconstruction Z. Zhao, T. Wang, S. Xia and Y. Wang This paper proposes a new system named as Hand-3D-Studio to capture the 3D hand pose and shape information.
498 Learning Endmember Dynamics in Multitemporal Hyperspectral Data Using A State-Space Model Formulation L. Drumetz, M. D. Mura, G. Tochon and R. Fablet In this paper, we propose a new framework for multitemporal unmixing and endmember extraction based on a state-space model, and present a proof of concept on simulated data to show how this representation can be used to inform multitemporal unmixing with external prior knowledge, or on the contrary to learn the dynamics of the quantities involved from data using neural network architectures adapted to the identification of dynamical systems.
499 Learning Eating Environments Through Scene Clustering S. K. Yarlagadda, S. Baireddy, D. G?era, C. J. Boushey, D. A. Kerr and F. Zhu In this paper, we propose an image clustering method to automatically extract the eating environments from eating occasion images captured during a community dwelling dietary study.
500 A Hybrid Structural Sparse Error Model for Image Deblocking Z. Zha, X. Yuan, J. Zhou, C. Zhu and B. Wen In this paper, we propose a novel hybrid structural sparse error (HSSE) model for image deblocking.
501 Reflectance-Guided, Contrast-Accumulated Histogram Equalization X. Wu, T. Kawanishi and K. Kashino To address this problem, we propose a histogram equalization-based method that adapts to the data-dependent requirements of brightness enhancement and improves the visibility of details without losing the global contrast.
502 Bilateral Recurrent Network for Single Image Deraining W. Shang, P. Zhu, D. Ren and H. Shi To address this issue, we in this paper propose bilateral recurrent network (BRN) to simultaneously exploit rain streak layer and background image layer.
503 Srzoo: An Integrated Repository For Super-Resolution Using Deep Learning J. Choi, J. Kim and J. Lee In this paper, we propose an integrated repository for the super-resolution tasks, named SRZoo, to provide state-of-the-art super-resolution models in a single place.
504 Sub-Dip: Optimization On A Subspace With Deep Image Prior Regularization And Application To Superresolution A. Sagel, A. Roumy and C. Guillemot This work proposes an alternative approach that relaxes this constraint and fully exploits all prior knowledge.
505 Parsing Map Guided Multi-Scale Attention Network For Face Hallucination C. Wang, Z. Zhong, J. Jiang, D. Zhai and X. Liu In this paper, we propose an effective two- step face hallucination method based on a deep neural network with multi-scale channel and spatial attention mechanism.
506 A Variational Bayesian Approach for Multichannel Through-Wall Radar Imaging with Low-Rank and Sparse Priors V. H. Tang, A. Bouzerdoum and S. L. Phung This paper considers the problem of multichannel through-wall radar (TWR) imaging from a probabilistic Bayesian perspective.
507 Semanticgan: Generative Adversarial Networks For Semantic Image To Photo-Realistic Image Translation J. Liu, Y. Zou and D. Yang To address those problems, we propose a SemanticGAN to synthesize high resolution image with fine details and realistic textures from the semantic label map.
508 Learning Blind Denoising Network for Noisy Image Deblurring M. Chen, Y. Chang, S. Cao and L. Yan In this work, we discover that the noise level and the denoiser is tightly coupled.
509 Pixel-Level Self-Paced Learning For Super-Resolution W. Lin, J. Gao, Q. Wang and X. Li To tackle this problem, this paper designs a training strategy named Pixel-level Self-Paced Learning (PSPL) to accelerate the convergence velocity of SISR models.
510 A Recursive Edge Detector For Color Filter Array Image B. Magnier, A. Aberkane and N. Gorrity In this paper, a new edge detection method is proposed for the computation of partial derivative images.
511 Image De-Raining Via RDL: When Reweighted Convolutional Sparse Coding Meets Deep Learning J. He, L. Yu and W. Yang Specifically, motivated by the success of reweighting algorithms, we propose solving the CSC model by learning weighted iterative soft thresholding algorithm (LwISTA) in a convolutional manner where the reweighted l1-norm is introduced.
512 CS-R-FCN: Cross-Supervised Learning for Large-Scale Object Detection Y. Guo, Y. Li and S. Wang In this paper, we present a novel cross-supervised learning pipeline for large-scale object detection, denoted as CS-R-FCN.
513 Dilated Convolutional Neural Networks for Panoramic Image Saliency Prediction F. Dai, Y. Zhang, Y. Ma, H. Li and Q. Zhao In this paper, we propose an encoder-decoder network for panoramic image saliency prediction.
514 Fine-Grained Action Recognition on a Novel Basketball Dataset X. Gu, X. Xue and F. Wang To tackle this issue, in this paper, we release a challenging dataset by annotating the fine-grained actions in basketball game videos.
515 Attention Guided Region Division for Crowd Counting X. Pan, H. Mo, Z. Zhou and W. Wu In this paper, we propose a two-branch network combining regression and detection.
516 Superpixel Segmentation Via Convolutional Neural Networks with Regularized Information Maximization T. Suzuki We propose an unsupervised superpixel segmentation method by optimizing a randomly-initialized convolutional neural network (CNN) in inference time.
517 Stacked Pooling for Boosting Scale Invariance of Crowd Counting S. Huang, X. Li, Z. Cheng, Z. Zhang and A. Hauptmann In this work, we take insight into the dense crowd counting problem by exploring the phenomenon of cross-scale visual similarity caused by perspective distortions.
518 GFNet: A Lightweight Group Frame Network for Efficient Human Action Recognition H. Liu, L. Zhang, L. Guan and M. Liu In this paper, we propose a lightweight neural network called Group Frame Network (GFNet) for human action recognition, which imposes intra-frame spatial information sparsity on spatial dimension in a simple yet effective way.
519 ROIMIX: Proposal-Fusion Among Multiple Images for Underwater Object Detection W. Lin, J. Zhong, S. Liu, T. Li and G. Li We propose an augmentation method called RoIMix, which characterizes interactions among images.
520 Tree of Shapes Cut for Material Segmentation Guided by a Design J. Baderot, M. Desvignes, L. Condat and M. D. Mura In this paper, we propose a method to segment different materials in a manufactured object.
521 Deep Flow Collaborative Network for Online Visual Tracking P. Liu, X. Yan, Y. Jiang and S. Xia In this paper, we propose an effective tracking algorithm to alleviate the time-consuming problem.
522 Salient Object Detection Based On Image Bit-Map B. Cao, X. Meng, S. Zhu and B. Zengv In this paper, we propose a novel salient object detection framework, which makes full use of the essential image compression.
523 A Novel Saliency-Driven Oil Tank Detection Method for Synthetic Aperture Radar Images L. Zhang and C. Liu In this paper, we propose a novel saliency-driven oil tank detection method (SDD) for SAR images.
524 Video Frame Interpolation Via Residue Refinement H. Li, Y. Yuan and Q. Wang In this paper, we propose a novel network structure that leverages residue refinement and adaptive weight to synthesize in-between frames.
525 Attention-Guided Deraining Network Via Stage-Wise Learning K. Jiang et al. To solve this problem, along with the stage-wise learning, we propose a novel attention-guided deraining network (ADN) for rain streak removal.
526 Attention-Mask Dense Merger (Attendense) Deep HDR for Ghost Removal K. Metwaly and V. Monga We propose a new deep HDR technique that does not need any explicit alignment of SDR images.
527 Y-Net: Multi-Scale Feature Aggregation Network With Wavelet Structure Similarity Loss Function For Single Image Dehazing H. Yang, C. H. Yang and Y. James Tsai In this paper, we propose a Y-net that is named for its structure.
528 Image Super-Resolution Using Residual Global Context Network K. Liu, Z. Han, J. Chen, C. Liu, J. Chen and Z. Wang To solve these problem,we introduce Global Context block (GCB) and design a comparative shallow network called Residual Global Context Networks (RGC-N).
529 Principle-Inspired Multi-Scale Aggregation Network for Extremely Low-Light Image Enhancement J. Zhang, R. Liu, L. Ma, W. Zhong, X. Fan and Z. Luo To address this issue, we develop a Principle-inspired Multi-scale Aggregation Network (PMA-Net) to simultaneously achieve the exposure enhancement and noises removal.
530 Non-Local Nested Residual Attention Network for Stereo Image Super-Resolution W. Xie, J. Zhang, Z. Lu, M. Cao and Y. Zhao To address this problem, in this paper, we propose a novel network named Non-local Nested Residual Attention Network (NNRANet).
531 Opendenoising: An Extensible Benchmark for Building Comparative Studies of Image Denoisers F. Lemarchand, E. F. Montesuma, M. Pelcat and E. Nogues This paper proposes a comparative study of existing denoisers, as well as an extensible open tool that makes it possible to reproduce and extend the study.
532 SDTCN: Similarity Driven Transmission Computing Network for Image Dehazing L. Zhang, S. Wang and X. Wang In this paper, we propose a novel light-weight similarity driven transmission computing network called SDTCN that is guided by the attributes of transmission similarity.
533 Joint Enhancement And Denoising of Low Light Images Via JND Transform L. Yu, H. Su and C. Jung In this paper, we propose joint enhancement and denoising of low light images via just-noticeable-difference (JND) transform.
534 Adversarial Text Image Super-Resolution using Sinkhorn Distance C. Geng, L. Chen, X. Zhang and Z. Gao In this paper, instead of using the Lp-norm as the supervision metric, we propose a novel one for better preserving semantic information in text images.
535 ADRN: Attention-Based Deep Residual Network for Hyperspectral Image Denoising Y. Zhao, D. Zhai, J. Jiang and X. Liu In this paper, we propose an attention-based deep residual network to directly learn a mapping from noisy HSI to the clean one.
536 Weakly Supervised Segmentation Guided Hand Pose Estimation During Interaction with Unknown Objects C. Zhang, G. Wang, X. Chen, P. Xie and T. Yamasaki To alleviate the influence of unknown objects, we propose a novel weakly supervised segmentation guided scheme to estimate hand poses.
537 Sparse Directed Graph Learning for Head Movement Prediction in 360 Video Streaming X. Zhang, G. Cheung, P. Le Callet and J. Z. G. Tan In this paper, we cast the head movement prediction task as a sparse directed graph learning problem, where three sources of relevant information?a 360 image saliency map, collected viewers? head movement traces, and a biological head rotation model?are aggregated into a unified Markov model.
538 Tracking to Improve Detection Quality in Lidar For Autonomous Driving J. Tang, A. Yellepeddi, S. Demirtas and C. Barber In this work, we leverage information from multiple consecutive frames to improve the detection capabilities of Lidar systems.
539 Cartoon-Texture Decomposition-Based Variational Pansharpening Y. Chen, M. Zhang, S. Li, Z. Wang and X. Tian In this paper, the similarities of MS and PAN images in cartoon-texture space are exploited.
540 An Alternative Signature Design Using L1 Principal Components for Spread-Spectrum Steganography C. P. Bailey, S. Chamadia and D. A. Pados This paper introduces novel spread spectrum (SS) and improved spread spectrum (ISS) multimedia data embedding techniques using L1 principal component signatures.
541 Privacy-Preserving Pattern Recognition Using Encrypted Sparse Representations in L0 Norm Minimization T. Nakachi, Y. Wang and H. Kiya In this paper, we propose a privacy-preserving pattern recognition method that uses encrypted sparse representations in L0 norm minimization.
542 Flexibly-Tunable Bitcube-Based Perceptual Encryption Within Jpeg Compression S. Shimizu and T. Suzuki We propose a perceptual encryption within JPEG compression (EWJ).
543 Gyroscope Aided Video Stabilization Using Nonlinear Regression on Special Orthogonal Group X. Hu, D. Olesen and P. Knudsen This paper presents a novel approach for gyroscope aided video stabilization.
544 Bband Index: a No-Reference Banding Artifact Predictor Z. Tu, J. Lin, Y. Wang, B. Adsumilli and A. C. Bovik Here we study this artifact, and propose a new distortion-specific no-reference video quality model for predicting banding artifacts, called the Blind BANding Detector (BBAND index).
545 Lqaid: Localized Quality Aware Image Denoising Using Deep Convolutional Neural Networks S. V. Reddy Dendi, C. Dev, N. Kothari and S. S. Channappayya In this paper we propose the Localized Quality Aware Image Denoising (LQAID) technique for image denoising using deep convolutional neural networks (CNNs).
546 Weakly Supervised Crowd-Wise Attention For Robust Crowd Counting X. Kong, M. Zhao, H. Zhou and C. Zhang In this paper, we propose a novel robust crowd counting method by introducing a weakly supervised crowd-wise attention network.
547 Xpsnr: A Low-Complexity Extension of The Perceptually Weighted Peak Signal-To-Noise Ratio For High-Resolution Video Quality Assessment C. R. Helmrich, M. Siekmann, S. Becker, S. Bosse, D. Marpe and T. Wiegand In this paper we show that, by way of low-complexity enhancements of our previous work on a perceptually weighted PSNR (WPSNR) metric, addressing shortcomings with video and ultra high-definition content, the prediction of human judgments of video coding quality by the WPSNR can be improved.
548 Non-Experts or Experts? Statistical Analyses of MOS using DSIS method Y. Sugito and M. Bertalm?o In this study, we analyze the results of several subjective evaluation experiments using the DSIS method.
549 Full Reference Video Quality Measures Improvement Using Neural Networks L. F. Tiotsop, A. Servetti and E. Masala To address such a situation, we propose a machine learning based improvement for each of the VQMs considered in this work and a video quality metric fusion index (VQMFI) that jointly exploits all the VQMs considered in the study as well as spatiotemporal features to produce a better estimation of the subjective quality.
550 Learning Multi-Scale Attentive Features for Series Photo Selection J. Huang, C. Cui, C. Zhang, Z. Shen, J. Yu and Y. Yin In this paper, we develop a novel deep CNN architecture that aggregates multi-scale features from different network layers, in order to capture the subtle differences between series photos.
551 Spatio-Temporal and Geometry Constrained Network for Automobile Visual Odometry H. Liu, P. Wei, W. Huang, G. Hua and F. Meng To overcome these deficiencies, an end-to-end framework that leverages spatio-temporal relevance and geometrical knowledge is proposed.
552 A Comprehensive Framework for 2D-JND Extension to 360-DEG Images S. Jaballah, A. Bhavsar and M. Larabi In this paper, a novel framework to extend 2D-JND models to estimate thresholds for 360-degree images is proposed.
553 Composite Dynamic Texture Synthesis Using Hierarchical Linear Dynamical System R. Singh, S. Yu and J. C. Principe We demonstrate that a systematic inclusion of prior structural constraints on the states of a linear dynamical system significantly improves its ability to model complex multidimensional sequences.
554 Steganography and its Detection in JPEG Images Obtained with the “TRUNC” Quantizer J. Butora and J. Fridrich Many portable imaging devices use the operation of “trunc” (rounding towards zero) instead of rounding as the final quantizer for computing DCT coefficients during JPEG compression. We show that this has rather profound consequences for steganography and its detection.
555 JPEG Steganography with Side Information from the Processing Pipeline Q. Giboulot, R. Cogranne and P. Bas In this paper, we propose a method to better estimate the variances of DCT coefficients by taking into account the dependencies between pixels that come from the development pipeline.
556 Selection-Channel-Aware Reverse JPEG Compatibility for Highly Reliable Steganalysis of JPEG Images R. Cogranne This paper deeply studies the principle of the recent reverse JPEG compatibility attack [1].
557 EMET: Embeddings from Multilingual-Encoder Transformer for Fake News Detection S. Schwarz, A. The?philo and A. Rocha This paper proposes an end-to-end framework called EMET to classify the reliability of small messages posted on social media platforms.
558 Depth Map Fingerprinting and Splicing Detection F. Matern, C. Riess and M. Stamminger In this work, we propose to use characteristic artifacts of depth reconstruction algorithms as trace for forensic analysis.
559 A Framework for Parameters Estimation of Image Operator Chain X. Liao and Z. Huang In this paper, we propose a new method to estimate the parameters of operations in different manipulation chains.
560 Privacy-Preserving Phishing Web Page Classification Via Fully Homomorphic Encryption E. J. Chou, A. Gururajan, K. Laine, N. K. Goel, A. Bertiger and J. W. Stokes This work introduces a fast and lightweight homomorphic-encryption pipeline that enables privacy-preserving machine learning for phishing web page recognition.
561 Privacy-Preserving Image Sharing Via Sparsifying Layers on Convolutional Groups S. Ferdowsi, B. Razeghi, T. Holotyak, F. P. Calmon and S. Voloshynovskiy We propose a practical framework to address the problem of privacy-aware image sharing in large-scale setups.
562 Evaluating Voice Conversion-Based Privacy Protection against Informed Attackers B. M. Lal Srivastava, N. Vauquier, M. Sahidullah, A. Bellet, M. Tommasi and E. Vincent In this paper, we investigate anonymization methods based on voice conversion.
563 Low-Complexity and Reliable Transforms for Physical Unclonable Functions O. G?nl? and R. F. Schaefer A new set of low-complexity and orthogonal transforms with no multiplication is proposed to obtain bit-error probability results significantly better than all methods previously proposed for key binding with PUFs.
564 Adversarial Detection of Counterfeited Printable Graphical Codes: Towards “Adversarial Games” In Physical World O. Taran, S. Bonev, T. Holotyak and S. Voloshynovskiy This paper addresses a problem of anti-counterfeiting of physical objects and aims at investigating a possibility of counterfeited printable graphical code detection from a machine learning perspectives.
565 Phylogenetic Minimum Spanning Tree Reconstruction Using Autoencoders R. Castelletto, S. Milani and P. Bestagini This paper proposes a matrix denoising solution that both mitigates dissimilarity noise and reconstruct the desired phylogenetic tree at the same time,
566 FCEM: A Novel Fast Correlation Extract Model For Real Time Steganalysis Of VoIP Stream Via Multi-Head Attention H. Yang, Z. Yang, Y. Bao, S. Liu and Y. Huang In this paper, we utilized attention mechanisms, which have recently attracted enormous interests due to their highly parallelizable computation and flexibility in modeling correlation in sequence, to tackle steganalysis problem of Quantization Index Modulation (QIM) based steganography in compressed VoIP stream.
567 Approaching Optimal Embedding In Audio Steganography With GAN J. Yang, H. Zheng, X. Kang and Y. Shi In this work, we proposed a framework based on Generative Adversarial Network (GAN) to approach optimal embedding for audio steganography in the temporal domain.
568 Multi-Stage Residual Hiding for Image-Into-Audio Steganography W. Cui, S. Liu, F. Jiang, Y. Liu and D. Zhao In this paper, we present a cross-modal steganography method for hiding image content into audio carriers while preserving the perceptual fidelity of the cover audio.
569 Patch-Level Selection and Breadth-First Prediction Strategy for Reversible Data Hiding H. Wu It motivates us to introduce a novel patch-level selection and breadth-first prediction strategy for efficient reversible data hiding.
570 Digital Watermarking For Protecting Audio Classification Datasets W. Kim and K. Lee In this study, we investigate the possibility of protecting audio classification datasets used in deep learning by embedding a pattern in the magnitude of the time-frequency representation of a subset of the dataset.
571 Saliency-Based Image Contrast Enhancement with Reversible Data Hiding S. Yang In this paper, a new approach is proposed by introducing the visual saliency into the process of reversible data hiding.
572 Unseen Face Presentation Attack Detection with Hypersphere Loss Z. Li, H. Li, K. Lam and A. C. Kot In this paper, we formulate the face presentation attack detection task under an open-set setting and address with our proposed deep anomaly detection based method.
573 Texception: A Character/Word-Level Deep Learning Model for Phishing URL Detection F. Tajaddodianfar, J. W. Stokes and A. Gururajan In this work, we propose a novel deep learning architecture, Texception, that takes a URL as input and predicts whether it belongs to a phishing attack.
574 Cell-Phone Classification: A Convolutional Neural Network Approach Exploiting Electromagnetic Emanations B. B. Yilmaz, E. Mert Ugurlu, A. Zajic and M. Prvulovic In this paper, we propose a methodology to identify both the brand of a cell-phone, and the status of its camera by exploiting electromagnetic (EM) emanations.
575 Quality-of-Service Prediction for Physical-layer Security via Secrecy Maps M. A. Gutierrez-Estevez et al. Building upon this concept, in this work we focus on system design-related aspects which consider physical-layer security as a service, together with thereby associated secrecy-related Quality-of-Service (QoS) requirements.
576 Secure Identification for Gaussian Channels W. Labidi, C. Deppe and H. Boche We focus on Gaussian channels for their known practical relevance.
577 Anti-Jamming Routing For Internet of Satellites: a Reinforcement Learning Approach C. Han, A. Liu, L. Huo, H. Wang and X. Liang This paper investigates anti-jamming routing scheme for heterogeneous IoS, with the aim of minimizing anti-jamming routing cost.
578 Electro-Magnetic Side-Channel Attack Through Learned Denoising and Classification F. Lemarchand, C. Marlin, F. Montreuil, E. Nogues and M. Pelcat This paper proposes an upgraded Electro Magnetic (EM) sidechannel attack that automatically reconstructs the intercepted data.
579 Detection of Malicious Vbscript Using Static and Dynamic Analysis with Recurrent Deep Learning J. W. Stokest, R. Agrawal and G. McDonald In this study, we explore a system that employs both static and dynamic analysis to detect malicious VBScripts.
580 Dynamic Attack Scoring Using Distributed Local Detectors Z. Zohrevand and U. Gl?sser This paper presents a novel study on detecting cyberattacks against distributed supervisory control systems.
581 Hijacking Tracker: A Powerful Adversarial Attack on Visual Tracking X. Yan, X. Chen, Y. Jiang, S. Xia, Y. Zhao and F. Zheng In this paper, we propose to add slight adversarial perturbations to the input image by an inconspicuous but powerful attack strategy?hijacking algorithm.
582 AdvMS: A Multi-Source Multi-Cost Defense Against Adversarial Attacks X. Wang, S. Wang, P. Chen, X. Lin and P. Chin In this paper, we study principles of designing multi-source and multi-cost schemes where defense performance is boosted from multiple defending components.
583 Classifying Anomalies for Network Security E. H. Do and V. N. Gadepally This work outlines a machine learning technique that uses deep neural networks to detect and classify a variety of network attacks.
584 A Switching Transmission Game with Latency as the User?s Communication Utility A. Garnaev, A. Petropulu, W. Trappe and H. V. Poor We consider the communication between a source (user) and a destination in the presence of a jammer, and study resource assignment in a non-cooperative game theory framework using communication latency as the user?s utility.
585 An Efficient Methodology to De-Anonymize the 5G-New Radio Physical Downlink Control Channel B. E. Gardner and J. D. Roth This paper presents an efficient approach to negating current anonymization methods in a computationally efficient manner.
586 Joint Learning of Assignment and Representation for Biometric Group Membership M. Gheisari, T. Furon and L. Amsaleg This paper proposes a framework for group membership protocols preventing the curious but honest server from reconstructing the enrolled biometric signatures and inferring the identity of querying clients.
587 Private FL-GAN: Differential Privacy Synthetic Data Generation Based on Federated Learning B. Xin, W. Yang, Y. Geng, S. Chen, S. Wang and L. Huang To address this issue, we propose private FL-GAN, a differential privacy generative adversarial network model based on federated learning.
588 Augmentation Data Synthesis Via Gans: Boosting Latent Fingerprint Reconstruction Y. Xu, Y. Wang, J. Liang and Y. Jiang To address these challenges, we propose a novel generative adversarial network (GAN) based data augmentation scheme to improve such reconstruction.
589 Learning to Fool the Speaker Recognition J. Li et al. In this paper, we attempt to fool the state-of-the-art speaker recognition model and present speaker recognition attacker, a lightweight model to fool the deep speaker recognition model by adding imperceptible perturbations onto the raw speech waveform.
590 Ts-Fen: Probing Feature Selection Strategy for Face Anti-Spoofing D. Peng, J. Xiao, R. Zhu and G. Gao In this paper, we propose a novel Two-Stream Feature Extraction Network (TS-FEN) based on depth and chrominance cues, guiding both sparsity and density of the feature distribution.
591 Improving Cross-Dataset Performance of Face Presentation Attack Detection Systems Using Face Recognition Datasets A. Mohammadi, S. Bhattacharjee and S. Marcel Here, we propose a novel PAD method that leverages the large variability present in FR datasets to induce invariance to factors that cause domain-shift.
592 SSTNet: Detecting Manipulated Faces Through Spatial, Steganalysis and Temporal Features X. Wu, Z. Xie, Y. Gao and Y. Xiao In this work, we propose a novel manipulation detection framework, named SSTNet, which detects tampered faces through Spatial, Steganalysis and Temporal features.
593 Multimodal Violence Detection in Videos B. Peixoto, B. Lavi, P. Bestagini, Z. Dias and A. Rocha In this study, we propose to analyze different concepts related to violence and how to better describe these concepts exploring visual and auditory cues in order to reach a robust method to detect violence.
594 Open Set Video Camera Model Verification O. Mayer, B. Hosler and M. C. Stamm In this work we propose a new, video-specific system for open set verification of camera models. We introduce a new open set video forensics problem called video camera model verification.
595 Multi-Patch Aggregation Models for Resampling Detection M. Lamba and K. Mitra To handle this issue, we propose two novel deep neural networks ? Iterative Pooling Network (IPN), which does not assume any prior information about the original image size, and Branched Network (BN), which uses this prior knowledge to produce better results.
596 Improving the Chronological Sorting of Images through Occlusion: A Study on the Notre-Dame Cathedral Fire R. Padilha, F. A. Andal? and A. Rocha In this work, we train a data-driven method to chronologically sort images originated from a real event, the Notre-Dame Cathedral fire, which broke out on April 15th, 2019.
597 Effectiveness of Random Deep Feature Selection for Securing Image Manipulation Detectors Against Adversarial Examples M. Barni, E. Nowroozi, B. Tondi and B. Zhang We investigate if the random feature selection approach proposed in [1] to improve the robustness of forensic detectors to targeted attacks, can be extended to detectors based on deep learning features.
598 A Dense U-Net with Cross-Layer Intersection for Detection and Localization of Image Forgery R. Zhang and J. Ni In this paper, we apply cross-layer intersection mechanism to dense u-net for image forgery detection and localization.
599 Fast Start-Up Algorithm for Adaptive Noise Cancellers with Novel SNR Estimation and Stepsize Control A. Sugiyama This paper proposes a fast convergence algorithm for adaptive noise cancellers with novel SNR (signal-to-noise ratio) estimation and stepsize control.
600 Robust and Computationally-Efficient Anomaly Detection Using Powers-Of-Two Networks U. Muneeb, E. Koyuncu, Y. Keshtkarjahromd, H. Seferoglu, M. F. Erden and A. Enis Cetin We propose a technique to increase robustness and reduce computational complexity in a Convolutional Neural Network (CNN) based anomaly detector that utilizes the optical flow information of video data.
601 Semi-Supervised Optimal Transport Methods for Detecting Anomalies A. Alaoui-Belghiti, S. Chevallier, E. Monacelli, G. Bao and E. Azabou Building upon advances on optimal transport and anomaly detection, we propose a generalization of an unsupervised and automatic method for detection of significant deviation from reference signals.
602 Learning to Estimate Driver Drowsiness from Car Acceleration Sensors Using Weakly Labeled Data T. Katsuki, K. Zhao and T. Yoshizumi By assuming that some aspects of driver drowsiness increase over time due to tiredness, we formulate an algorithm that can learn from such weakly labeled data.
603 Damage-Sensitive and Domain-Invariant Feature Extraction for Vehicle-Vibration-Based Bridge Health Monitoring J. Liu, B. Chen, S. Chen, M. Berg?s, J. Bielak and H. Noh We introduce a physics-guided signal processing approach to extract a damage-sensitive and domain-invariant (DS & DI) feature from acceleration response data of a vehicle traveling over a bridge to assess bridge health.
604 On Robust Variance Filtering and Change of Variance Detection Q. Wen, Z. Ma and L. Sun To deal with these challenges, we propose a robust CoV detection algorithm based on robust statistics and sparse regularizations.
605 Low-Rank Gradient Approximation for Memory-Efficient on-Device Training of Deep Neural Network M. Gooneratne, K. C. Sim, P. Zadrazil, A. Kabel, F. Beaufays and G. Motta In this paper, we propose approximating the gradient matrices of deep neural networks using a low-rank parameterization as an avenue to save training memory.
606 Improving Efficiency in Large-Scale Decentralized Distributed Training W. Zhang et al. In this paper, we investigate techniques to accelerate (A)D-PSGD based training by improving the spectral gap while minimizing the communication cost.
607 Parallelizing Adam Optimizer with Blockwise Model-Update Filtering K. Chen, H. Ding and Q. Huo In this paper, we attempt to parallelize Adam with blockwise model-update filtering (BMUF) instead.
608 Joint Training of Deep Neural Networks for Multi-Channel Dereverberation and Speech Source Separation M. Togami In this paper, we propose a joint training of two deep neural networks (DNNs) for dereverberation and speech source separation.
609 Structural Sparsification for Far-Field Speaker Recognition with Intel? Gna J. Zhang, J. Huang, M. Deisher, H. Li and Y. Chen In this paper, we apply structural sparsification on time-delay neural networks (TDNN) to remove redundant structures and accelerate the execution.
610 Environment-Aware Reconfigurable Noise Suppression J. Yang and J. Bingham The paper proposes an efficient, robust, and reconfigurable technique to suppress various types of noises for any sampling rate.
611 Fully-Neural Approach to Heavy Vehicle Detection on Bridges Using a Single Strain Sensor T. Kawakatsu, K. Aihara, A. Takasu and J. Adachi In this paper, we propose a novel BWIM mechanism, which employs a deep convolutional neural network (CNN).
612 Multichannel Signal Processing for Road Surface Identification G. Safont, A. Salazar, A. Rodriguez and L. Vergara This work introduces a multi-sensor road surface identification system that considers features from four different kind of sensors: microphones, accelerometers, speed signals, and handwheel signals.
613 A Monte Carlo Search-Based Triplet Sampling Method for Learning Disentangled Representation of Impulsive Noise on Steering Gear S. Bu, N. Park, G. Nam, J. Seo and S. Cho In this paper, we propose a method to overcome the above two major hurdles by modify a sampling algorithm of triplet pairs based on structural similarity index instead of naive Euclidean distance within Monte Carlo based sampling strategy.
614 Stochastic Geometry Planning of Electric Vehicles Charging Stations R. Atat, M. Ismail and E. Serpedin In this paper, using stochastic geometry, we formulate the CSs planning on a stochastic geometry-based power grid model, that we previously showed to mimic real-world power grids.
615 Discriminant Generative Adversarial Networks with its Application to Equipment Health Classification S. Zheng and C. Gupta We propose discriminant GANs, where, for generated samples, we maximize between-class distance and minimize within-class distance, so that generated samples in different classes are more separable and different betweenclass distances are explicitly allowed.
616 Power Optimization Using Embedded Automatic Gain Control Algorithm with Photoplethysmography Signal Quality Classification F. Foroozan, D. D. Xue, K. Fang and J. S. James Wu This paper presents the design and implementation of an Automatic Gain Control (AGC) embedded algorithm for photoplethysmographic (PPG) sensors.
617 A General Difficulty Control Algorithm for Proof-of-Work Based Blockchains S. Zhang and X. Ma This paper proposes a general difficulty control algorithm and provides insights for difficulty adjustment rules for PoW based blockchains.
618 A New Application of Ultrasound Signal Processing for Archaeological Ceramic Classification A. Salazar, G. Safont and L. Vergara This work proposes a novel system that overcomes those limitations using non-destructive ultrasonic testing and advanced pattern recognition techniques.
619 Headless Horseman: Adversarial Attacks on Transfer Learning Models A. Abdelkader et al. We present a family of transferable adversarial attacks against such classifiers, generated without access to the classification head; we call these headless attacks.
620 Detecting Adversarial Attacks In Time-Series Data M. G. Abdu-Aguye, W. Gomaa, Y. Makihara and Y. Yagi We propose a method for detecting Fast Gradient Sign Method (FGSM) and Basic Iterative Method (BIM) adversarial attacks as adapted for time-series data.
621 Detection of Adversarial Attacks and Characterization of Adversarial Subspace M. Esmaeilpour, P. Cardinal and A. L. Koerich In this paper, we explore subspaces of adversarial examples in unitary vector domain, and we propose a novel detector for defending our models trained for environmental sound classification.
622 Adversarial Example Detection by Classification for Deep Speech Recognition S. Samizade, Z. Tan, C. Shen and X. Guan We, however, formulate the defense as a classification problem and present a strategy for systematically generating adversarial example datasets: one for white-box attacks and one for black-box attacks, containing both adversarial and normal examples.
623 Characterizing Speech Adversarial Examples Using Self-Attention U-Net Enhancement C. Yang, J. Qi, P. Chen, X. Ma and C. Lee In this work, we present a U-Net based attention model, UNetAt, to enhance adversarial speech signals.
624 Action-Manipulation Attacks on Stochastic Bandits G. Liu and L. Lai In this paper, we propose a new class of attack named action-manipulation attack, where an adversary can change the action signal selected by the user.
625 Primal-Dual Stochastic Subgradient Method For Log-Determinant Optimization S. Wu, H. Yu and J. Dauwels This paper proposes a highly efficient stochastic method that has time complexity $\mathcal{O}$(N2), whereas existing methods have complexity $\mathcal{O}$(N3).
626 Neural Network Training with Approximate Logarithmic Computations A. Sanyal, P. A. Beerel and K. M. Chugg This paper proposed an end-to-end training and inference scheme that eliminates multiplications by approximate operations in the log-domain which has the potential to significantly reduce implementation complexity.
627 Automatic and Simultaneous Adjustment of Learning Rate and Momentum for Stochastic Gradient-based Optimization Methods T. Lancewicki and S. Kopru In this paper, we propose a generic approach that utilizes the statistics of an unbiased gradient estimator to automatically and simultaneously adjust two paramount hyperparameters: the learning rate and momentum.
628 A Study of Generalization of Stochastic Mirror Descent Algorithms on Overparameterized Nonlinear Models N. Azizan, S. Lale and B. Hassibi We study the convergence, the implicit regularization and the generalization of stochastic mirror descent (SMD) algorithms in overparameterized nonlinear models, where the number of model parameters exceeds the number of training data points.
629 On Distributed Stochastic Gradient Descent for Nonconvex Functions in the Presence of Byzantines S. Bulusu, P. Khanduri, P. Sharma and P. K. Varshney We propose a robust variant of distributed SGD which is resilient to the presence of Byzantines.
630 Preconditioning ADMM for Fast Decentralized Optimization M. Ma and G. B. Giannakis In this work, we consider the distributed optimization problem using networked computing machines.
631 Extrapolated Alternating Algorithms for Approximate Canonical Polyadic Decomposition A. M. Shun Ang, J. E. Cohen, L. T. Khanh Hien and N. Gillis In this work, we propose several algorithms based on extrapolation that improve over existing alternating methods for aCPD.
632 Scalable Kernel Learning Via the Discriminant Information M. Al, Z. Hou and S. Kung By generalizing this measure to cover a wider range of kernel maps and learning settings, we develop scalable methods to learn kernel features with high discriminant power.
633 Arsm Gradient Estimator for Supervised Learning to Rank S. Z. Dadaneh, S. Boluki, M. Zhou and X. Qian We propose a new model for supervised learning to rank.
634 Solving Non-Convex Non-Differentiable Min-Max Games Using Proximal Gradient Method B. Barazandeh and M. Razaviyayn In this work, we study special form of non-smooth min-max games when the objective function is (strongly) convex with respect to one of the player?s decision variable.
635 A Fast and Accurate Frequent Directions Algorithm for Low Rank Approximation via Block Krylov Iteration Q. Yi, C. Wang, X. Liao and Y. Wang Specifically, by utilizing the power of Block Krylov Iteration and count sketch techniques, we propose a fast and accurate FD algorithm dubbed as BKICS-FD.
636 Stochastic Admm For Byzantine-Robust Distributed Learning F. Lin, Q. Ling, W. Li and Z. Xiong In this paper, we aim at solving a distributed machine learning problem under Byzantine attacks.
637 Unified Signal Compression Using Generative Adversarial Networks B. Liu, A. Cao and H. Kim We propose a unified compression framework that uses generative adversarial networks (GAN) to compress image and speech signals.
638 Wind: Wasserstein Inception Distance For Evaluating Generative Adversarial Network Performance P. Dimitrakopoulos, G. Sfikas and C. Nikou In this paper, we present Wasserstein Inception Distance (WInD), a novel metric for evaluating performance of Generative Adversarial Networks (GANs).
639 Trace Norm Generative Adversarial Networks for Sensor Generation and Feature Extraction S. Zheng and C. Gupta In this work, we propose Trace Norm GANs to enforce trace norm minimization on generated samples.
640 Mahalanobis Distance Based Adversarial Network for Anomaly Detection Y. Hou, Z. Chen, M. Wu, C. Foo, X. Li and R. M. Shubair This paper proposes an efficient method, known as Mahalanobis Distance-based Adversarial Network (MDAN), for anomaly detection.
641 Commuting Conditional GANS for Multi-Modal Fusion S. Roheda, H. Krim and B. S. Riggan This paper presents a data driven approach to multi-modal fusion where a hidden latent sub-space between the different modalities is learned.
642 Sequence-To-Subsequence Learning With Conditional Gan For Power Disaggregation Y. Pan, K. Liu, Z. Shen, X. Cai and Z. Jia In this paper, we propose a sequence-to-subsequence learning method, which makes a trade-off between traditional sequence-to-sequence and sequence-to-point method, to balance the convergence difficulty in deep neural networks and the amount of computation in the inference period.
643 A Bin Encoding Training of a Spiking Neural Network Based Voice Activity Detection G. Dellaferrera, F. Martinelli and M. Cernak In this work, we develop a SNN-based Voice Activity Detection (VAD) system that belongs to the building blocks of any audio and speech processing system.
644 ECG Heartbeat Classification Based on Multi-Scale Wavelet Convolutional Neural Networks L. E. Bouny, M. Khalil and A. Adib This paper proposes a novel Deep Learning technique for ECG beats classification.
645 Self-Supervised Learning for ECG-Based Emotion Recognition P. Sarkar and A. Etemad We present an electrocardiogram (ECG) -based emotion recognition system using self-supervised learning.
646 Expression-Guided EEG Representation Learning for Emotion Recognition S. Rayatdoost, D. Rudrauf and M. Soleymani In this paper, we propose a deep representation learning approach for emotion recognition from electroencephalogram (EEG) signals guided by facial electromyogram (EMG) and electrooculogram (EOG) signals.
647 Attention Driven Fusion for Multi-Modal Emotion Recognition D. Priyasad, T. Fernando, S. Denman, S. Sridharan and C. Fookes In this paper, we present a deep learning-based approach to exploit and fuse text and acoustic data for emotion classification.
648 Learning the Spatio-Temporal Dynamics of Physical Processes from Partial Observations I. Ayed, E. d. B?zenac, A. Pajot and P. Gallinari We propose a data-driven framework, where the system?s dynamics are modeled by an unknown time-varying differential equation and the evolution term for the state is estimated from the partially observed data only, using a deep convolutional neural network.
649 Optimal Laplacian Regularization for Sparse Spectral Community Detection L. Dall?Amico, R. Couillet and N. Tremblay In this paper we formally determine a proper regularization which is intimately related to alternative state-of-the-art spectral techniques for sparse graphs.
650 Anomaly Detection in Mixed Time-Series Using A Convolutional Sparse Representation With Application To Spacecraft Health Monitoring B. Pilastre, G. Silva, L. Boussouf, S. d?Escrivan, P. Rodr?guez and J. Tourneret This paper introduces a convolutional sparse model for anomaly detection in mixed continuous and discrete data.
651 Variational Student: Learning Compact and Sparser Networks In Knowledge Distillation Framework S. Hegde, R. Prasad, R. Hebbalaguppe and V. Kumar To this end, we propose Variational Student where we reap the benefits of compressibility of the knowledge distillation framework, and sparsity inducing abilities of variational inference (VI) techniques.
652 Low Rank Activations for Tensor-Based Convolutional Sparse Coding P. Humbert, J. Audiffren, L. Oudre and N. Vayatis In this article, we propose to extend the classical Convolutional Sparse Coding model (CSC) to multivariate data by introducing a new tensor CSC model that enforces sparsity and low-rank constraint on the activations.
653 Energy Disaggregation Using Fractional Calculus P. A. Schirmer and I. Mporas In this article we introduce the use of fractional calculus in the Non-Intrusive Load Monitoring task.
654 Dyna-Bolt: Domain Adaptive Binary Factorization Of Current Waveforms For Energy Disaggregation B. Chen, J. Liu, H. Lange and M. Berg?s To alleviate this problem, we formulated NILM in a domain adaptation context. Using Binary OnLine FactorizaTion (BOLT) as the baseline model, we first demonstrate that direct application of domain adversarial training without application-specific modifications is unsuccessful, and hypothesize that this may be due to differences in the data distributions.
655 Unsupervised Style and Content Separation by Minimizing Mutual Information for Speech Synthesis T. Hu, A. Shrivastava, O. Tuzel and C. Dhir We present a method to generate speech from input text and a style vector that is extracted from a reference speech signal in an unsupervised manner, i.e., no style annotation, such as speaker information, is required.
656 Improving Singing Voice Separation with the Wave-U-Net Using Minimum Hyperspherical Energy J. Perez-Lapillo, O. Galkin and T. Weyde In this work, we apply MHE regularization to the 1D filters of the Wave-U-Net.
657 Singing Voice Conversion with Disentangled Representations of Singer and Vocal Technique Using Variational Autoencoders Y. Luo, C. Hsu, K. Agres and D. Herremans We propose a flexible framework that deals with both singer conversion and singers vocal technique conversion.
658 Online Tensor Completion and Free Submodule Tracking With The T-SVD K. Gilman and L. Balzano We propose a new online algorithm, called TOUCAN, for the tensor completion problem of imputing missing entries of a low tubal-rank tensor using the tensor-tensor product (t- product) and tensor singular value decomposition (t-SVD) algebraic framework.
659 Exploiting Commutativity Condition for CP Decomposition Via Approximate Simultaneous Diagonalization R. Akema, M. Yamagishi and I. Yamada In this paper, we propose a novel strategy which utilizes an inherent algebraic property of simultaneously diagonalizable matrix tuples, i.e., commutativity, for both (i) reducing approximate CP decomposition of a higher-order tensor to Approximate Simultaneous Diagonalization (ASD) and (ii) solving the ASD.
660 A Novel Rank Selection Scheme in Tensor Ring Decomposition Based on Reinforcement Learning for Deep Neural Networks Z. Cheng, B. Li, Y. Fan and Y. Bao To overcome this issue, we propose a novel rank selection scheme, which is inspired by reinforcement learning, to automatically select ranks in recently studied tensor ring decomposition in each convolutional layer.
661 Estimating Structural Missing Values Via Low-Tubal-Rank Tensor Completion H. Wang, F. Zhang, J. Wang and Y. Wang In this paper, we propose to incorporate a constraint item on the missing values into low-tubal-rank tensor completion to promote the structural hypothesis of the missing values such as sparsity.
662 Low-Tubal-Rank Tensor Recovery From One-Bit Measurements J. Hou, F. Zhang, Y. Wang and J. Wang This paper focuses on the recovery of low-tubal-rank tensors from binary measurements under the frame of tensor Singular Value Decomposition.
663 Continual Learning Through One-Class Classification Using VAE F. Wiewel, A. Brendle and B. Yang In this paper, we propose a new method for overcoming this phenomenon based on one-class classification.
664 Estimation of Post-Nonlinear Causal Models Using Autoencoding Structure K. Uemura and S. Shimizu In this paper, we proposed a new estimation method of PNL model using an autoencoding structure.
665 From Symbols to Signals: Symbolic Variational Autoencoders C. Devaraj, A. Chowdhury, A. Jain, J. R. Kubricht, P. Tu and A. Santamaria-Pang We introduce Symbolic Variational Autoencoders which generate images from symbols that represent semantic concepts.
666 Graph Auto-Encoder for Graph Signal Denoising T. H. Do, D. Minh Nguyen and N. Deligiannis To bridge this gap, we propose to use graph convolutional neural network with a Kron-reduction-based pooling operator for denoising on graphs.
667 A Priori Estimates of the Generalization Error for Autoencoders Z. Don, W. E and C. Ma In this paper, we build theoretical understanding about autoencoders.
668 GFCN: A New Graph Convolutional Network Based on Parallel Flows F. Ji, J. Yang, Q. Zhang and W. P. Tay In this paper, we study the problem from a different perspective, by introducing parallel flow decomposition of graphs.
669 Depthwise-STFT Based Separable Convolutional Neural Networks S. Kumawat and S. Raman In this paper, we propose a new convolutional layer called Depthwise-STFT Separable layer that can serve as an alternative to the standard depthwise separable convolutional layer.
670 Semi-Implicit Stochastic Recurrent Neural Networks E. Hajiramezanali, A. Hasanzadeh, N. Duffield, K. Narayanan, M. Zhou and X. Qian In this paper, we advocate learning implicit latent representations using semi-implicit variational inference to further increase model flexibility.
671 Feedback Recurrent Autoencoder Y. Yang, G. Sauti?re, J. J. Ryu and T. S. Cohen In this work, we propose a new recurrent autoencoder architecture, termed Feedback Recurrent AutoEncoder (FRAE), for online compression of sequential data with temporal dependency.
672 Indylstms: Independently Recurrent LSTMS P. Gonnet and T. Deselaers We introduce Independently Recurrent Long Short-term Memory cells: IndyLSTMs.
673 Neural Attentive Multiview Machines O. Barkan, O. Katz and N. Koenigstein To this end, we introduce NAM: a Neural Attentive Multiview machine that learns multiview item representations and similarity by employing a novel attention mechanism.
674 Attentive Modality Hopping Mechanism for Speech Emotion Recognition S. Yoon, S. Dey, H. Lee and K. Jung In this work, we explore the impact of visual modality in addition to speech and text for improving the accuracy of the emotion detection system.
675 Facial Emotion Recognition Using Light Field Images with Deep Attention-Based Bidirectional LSTM A. Sepas-Moghaddam, A. Etemad, F. Pereira and P. L. Correia This paper exploits the rich spatio-angular information available in light field images for facial emotion recognition.
676 A Regularized Attention Mechanism for Graph Attention Networks U. S. Shanthamallu, J. J. Thiagarajan and A. Spanias In this paper, we perform a detailed analysis of GAT models, and present interesting insights into their behavior.
677 Attentive Item2vec: Neural Attentive User Representations O. Barkan, A. Caciularu, O. Katz and N. Koenigstein In this work, we present Attentive Item2vec (AI2V) – a novel attentive version of Item2vec (I2V).
678 Audio Sound Determination Using Feature Space Attention Based Convolution Recurrent Neural Network X. Xia, J. Pan and Y. Wang To deal with this, we propose a feature space attention based convolution recurrent neural network approach utilizing the varying importance of each feature dimension to perform acoustic event detection.
679 Spatial Attentional Bilinear 3D Convolutional Network for Video-Based Autism Spectrum Disorder Detection K. Sun, L. Li, L. Li, N. He and J. Zhu In this paper, we propose spatial attentional bilinear pooling to enhance its spatial information extraction without significantly increasing the parameters.
680 Linear Thompson Sampling Under Unknown Linear Constraints A. Moradipari, M. Alizadeh and C. Thrampoulidis In this setting, we propose Safe-LTS, the first safe Thompson Sampling based algorithm, and we prove that it achieves no-regret learning.
681 Overlapped State Hidden Semi-Markov Model for Grouped Multiple Sequences H. Narimatsu and H. Kasai To tackle this challenge, we propose a new model designated as overlapped state hidden semi-Markov model (OS-HSMM).
682 Online Community Detection by Spectral Cusum M. Zhang, L. Xie and Y. Xie We present an online community change detection algorithm called spectral CUSUM to detect the emergence of a community using a subspace projection procedure based on a Gaus-sian model setting.
683 Enhanced Adversarial Strategically-Timed Attacks Against Deep Reinforcement Learning C. H. Yang et al. In this paper, we introduce timing-based adversarial strategies against a DRL-based navigation system by jamming in physical noise patterns on the selected time frames.
684 Preference-Aware Mask for Session-Based Recommendation with Bidirectional Transformer Y. Zhang et al. In this paper, we propose the preference-aware mask to capture user preferences over the items within the sessions, which adapts to the preference-irrelevant items within the sessions and provides explainable evidence for the recommendation.
685 Low Mutual and Average Coherence Dictionary Learning Using Convex Approximation J. Parsa, M. Sadeghi, M. Babaie-Zadeh and C. Jutten In this paper, we consider a dictionary learning problem regularized with the average coherence and constrained by an upper-bound on the mutual coherence of the dictionary.
686 Robust Online Matrix Completion with Gaussian Mixture Model C. Liu, C. Chen, H. Shan and B. Wang In this paper, we study the problem of online matrix completion (MC) aiming to achieve robustness to the variations in both low-rank subspace and noises.
687 Deep Neural Network Based Matrix Completion for Internet of Things Network Localization S. Kim, L. T. Nguyen and B. Shim In this paper, we propose a deep neural network based matrix completion approach for Internet of Things (IoT) localization.
688 Bringing in the Outliers: A Sparse Subspace Clustering Approach to Learn a Dictionary of Mouse Ultrasonic Vocalizations J. Wang, K. Mundnich, A. T. Knoll, P. Levitt and S. Narayanan We propose a new method to automatically create a dictionary of USVs based on a two-step spectral clustering approach, where we split the set of USVs into inlier and outlier data sets.
689 One-Bit Compressed Sensing Using Generative Models G. Joseph, S. Kafle and P. K. Varshney In this paper, we address the classical problem of one-bit compressed sensing.
690 Hybrid Deep-Semantic Matrix Factorization for Tag-Aware Personalized Recommendation Z. Xu, D. Yuan, T. Lukasiewicz, C. Chen, Y. Miao and G. Xu Therefore, inspired by the recent development of deep-semantic modeling, we propose a hybrid deep-semantic matrix factorization (HDMF) model to further improve the performance of tag-aware personalized recommendation by integrating the techniques of deep-semantic modeling, hybrid learning, and matrix factorization.
691 Supervised Encoding for Discrete Representation Learning C. P. Le, Y. Zhou, J. Ding and V. Tarokh In this paper, we propose a novel supervised learning model named Supervised-Encoding Quantizer (SEQ).
692 Learning Data Representation and Emotion Assessment from Physiological Data M. S. Joaquim et al. In this work, raw two-channel pre-frontal electroencephalography and photoplethysmography signals of 25 subjects were collected using EMOTAI?s headband while watching commercials.
693 Feature Selection Under Orthogonal Regression with Redundancy Minimizing X. Xu and X. Wu To address the defect, in this paper, we propose a two-stage (filter-embedded) feature selection technique based on Maximum Relevance Minimum Redundancy and FSOR, termed as Orthogonal Regression with Minimum Redundancy (ORMR).
694 The Picasso Algorithm for Bayesian Localization Via Paired Comparisons in a Union of Subspaces Model G. Canal et al. We develop a framework for localizing an unknown point w using paired comparisons of the form “w is closer to point x i than to x j ” when the points lie in a union of known subspaces.
695 Learning Semi-Supervised Anonymized Representations by Mutual Information C. Feutry, P. Piantanida and P. Duhamel This paper addresses the problem of removing from a set of data (here images) a given private information, while still allowing other utilities on the processed data.
696 Learning Local Structure of Representative Points for Point Cloud Classification and Semantic Segmentation X. Li, Y. Pang, Y. Wu and Y. Li To solve this problem, we propose a learning-based block, named Representative Points Block (RPB), to select the most representative points of an irregular point cloud according to the task.
697 Towards Blind Quality Assessment of Concert Audio Recordings Using Deep Neural Networks N. Simou, Y. Mastorakis and N. Stefanakis In this work, we apply different Deep Neural Network (DNN) architectures to a simple binary classification problem, that of deciding whether a musical recording is user-generated or of professional quality.
698 Multi-Label Sound Event Retrieval Using A Deep Learning-Based Siamese Structure With A Pairwise Presence Matrix J. Fan, E. Nichols, D. Tompkins, A. E. M?ndez M?ndez, B. Elizalde and P. Pasquier To address this latter problem, we propose different Deep Learning architectures with a Siamesestructure and a Pairwise Presence Matrix.
699 Speech-Driven Facial Animation Using Polynomial Fusion of Features T. Kefalas, K. Vougioukas, Y. Panagakis, S. Petridis, J. Kossaifi and M. Pantic In this paper we propose a polynomial fusion layer that models the joint representation of the encodings by a higher-order polynomial, with the parameters modelled by a tensor decomposition.
700 SED-MDD: Towards Sentence Dependent End-To-End Mispronunciation Detection and Diagnosis Y. Feng, G. Fu, Q. Chen and K. Chen In order to integrate these stages, we propose SED-MDD, an end-to-end model for sentence dependent mispronunciation detection and diagnosis (MD&D) .
701 Generative Pre-Training for Speech with Autoregressive Predictive Coding Y. Chung and J. Glass In this paper we propose to use autoregressive predictive coding (APC), a recently proposed self-supervised objective, as a generative pre-training approach for learning meaningful, non-specific, and transferable speech representations.
702 Stargan for Emotional Speech Conversion: Validated by Data Augmentation of End-To-End Emotion Recognition G. Rizos, A. Baird, M. Elliott and B. Schuller In this paper, we propose an adversarial network implementation for speech emotion conversion as a data augmentation method, validated by a multi-class speech affect recognition task.
703 Multimodal Transformer Fusion for Continuous Emotion Recognition J. Huang, J. Tao, B. Liu, Z. Lian and M. Niu In this work, we utilize the Transformer model to fuse audio-visual modalities on the model level.
704 HKA: A Hierarchical Knowledge Attention Mechanism for Multi-Turn Dialogue System J. Song, K. Zhang, X. Zhou and J. Wu Motivated by this, we propose a novel hierarchical knowledge attention (HKA) mechanism for open-domain multi-turn dialogue system in this paper, which utilizes both word and utterance level attention jointly.
705 Submodular Rank Aggregation on Score-Based Permutations for Distributed Automatic Speech Recognition J. Qi, C. H. Yang and J. Tejedor This work studies the use of submodular functions to design a rank aggregation on score-based permutations, which can be used for distributed ASR systems in both supervised and unsupervised modes.
706 Bridging Mixture Density Networks with Meta-Learning for Automatic Speaker Identification R. Li, J. Jiang, X. Wu, H. Mao, C. Hsieh and W. Wang To alleviate the disadvantage caused by training data deficiency, we propose a Mixture Density Network- based Meta-Learning method (MDNML) for speaker identification.
707 Pitch Estimation Via Self-Supervision B. Gfeller, C. Frank, D. Roblek, M. Sharifi, M. Tagliasacchi and M. Velimirovic We present a method to estimate the fundamental frequency in monophonic audio, often referred to as pitch estimation.
708 Hierarchical Caching via Deep Reinforcement Learning A. Sadeghi, G. Wang and G. B. Giannakis To model the two-way interactive influence between caching decisions at the parent and leaf nodes, a reinforcement learning formulation is put forth in this work.
709 Learning Network Representation Through Reinforcement Learning S. Shen et al. In contrast, we propose, reinforcement learning network representations (RLNet), which explores the idea of using reinforcement learning to learn to explore the network and to obtain network representations.
710 Attention-Based Curiosity-Driven Exploration in Deep Reinforcement Learning P. Reizinger and M. Szemenyei Combining them, we propose new methods, such as Attention-aided Advantage Actor-Critic, an extension of the Actor-Critic framework.
711 Stabilizing Multi-Agent Deep Reinforcement Learning by Implicitly Estimating Other Agents? Behaviors Y. Jin, S. Wei, J. Yuan, X. Zhang and C. Wang To cope with this challenge, we propose a novel approach where each agent uses an implicit estimate of others? actions to guide its own policy learning.
712 QOS-Aware Flow Control for Power-Efficient Data Center Networks with Deep Reinforcement Learning P. Sun, Z. Guo, S. Liu, J. Lan and Y. Hu In this paper, we propose SmartFCT, which employs Software-Defined Networking (SDN) coupled with the Deep Reinforcement Learning (DRL) to improve the power efficiency of DCNs and guarantee the FCT.
713 Improving the Scalability of Deep Reinforcement Learning-Based Routing with Control on Partial Nodes P. Sun, J. Lan, Z. Guo, Y. Xu and Y. Hu In this paper, we propose SINET, a scalable and intelligent network control framework for routing optimization.
714 Generalized Linear Bandits with Safety Constraints S. Amani, M. Alizadeh and C. Thrampoulidis This paper formulates a generalized linear stochastic multi-armed bandit problem with generalized linear safety constraints that depend on an unknown parameter vector.
715 From Video Game to Real Robot: The Transfer Between Action Spaces J. Karttunen, A. Kanervisto, V. Kyrki and V. Hautam?ki In this work, we study how general video games can be directly used instead of fine-tuned simulations for the sim-to-real transfer.
716 Correlated Multi-Armed Bandits with A Latent Random Source S. Gupta, G. Joshi and O. Yagan We propose and analyze the performance of the C-UCB algorithm that leverages the correlations between arms to reduce the cumulative regret (i.e., to increase the total reward obtained after T rounds).
717 Adaptive Sequential Interpolator Using Active Learning for Efficient Emulation of Complex Systems L. Martino, D. H. Svendsen, J. Vicent and G. Camps-Valls This paper introduces an interpolation procedure which belongs to the family of active learning algorithms, in order to construct cheap surrogate models of such costly complex systems.
718 Continual Learning for Infinite Hierarchical Change-Point Detection P. Moreno-Mu?oz, D. Ram?rez and A. Art?s-Rodr?guez To circumvent this problem, we propose to use a hierarchical model, which yields observations that belong to a lower-dimensional manifold.
719 Cost Aware Adversarial Learning S. De Silva, J. Kim and R. Raich In this paper, we present an attack-cost-aware adversarial learning framework that takes into account the (potentially inhomogeneous) vulnerability characteristics of test data entries in designing an attack-resilient classifier.
720 On Divergence Approximations for Unsupervised Training of Deep Denoisers Based on Stein?s Unbiased Risk Estimator S. Soltanayev, R. Giryes, S. Y. Chun and Y. C. Eldar In this work, we briefly study the computational efficiency of Monte-Carlo (MC) divergence approximation over recently available exact divergence computation using backpropagation.
721 Variable Metric Proximal Gradient Method with Diagonal Barzilai-Borwein Stepsize Y. Park, S. Dhar, S. Boyd and M. Shah This paper proposes an adaptive metric selection strategy called diagonal Barzilai-Borwein (DBB) stepsize for the popular Variable Metric Proximal Gradient (VM-PG) algorithm [1], [2].
722 Revisit of Estimate Sequence for Accelerated Gradient Methods B. Li, M. Couti?o and G. B. Giannakis In this paper, we revisit the problem of minimizing a convex function f(x) with Lipschitz continuous gradient via accelerated gradient methods (AGM).
723 A Generalization of Principal Component Analysis S. Battaglino and E. Koyuncu We present a gradient ascent algorithm to solve the problem.
724 An Easy-to-Implement Framework of Fast Subspace Clustering For Big Data Sets L. Meng, Y. Jiao and Y. Gu To enable the fast implementation of subspace clustering on big datasets, this paper proposes a simple but effective subspace clustering framework called Fast Subspace Clustering (FSC), which adopts a “sampling, random projecting, clustering, and classifying” strategy.
725 Investigating Generalization in Neural Networks Under Optimally Evolved Training Perturbations S. Chaudhury and T. Yamasaki In this paper, we study the generalization properties of neural networks under input perturbations and show that minimal training data corruption by a few pixel modifications can cause drastic overfitting.
726 Heterogeneous Domain Generalization Via Domain Mixup Y. Wang, H. Li and A. C. Kot In this work, we focus on the problem of heterogeneous domain generalization which aims to improve the generalization capability across different tasks, which is, how to learn a DCNN model with multiple domain data such that the trained feature extractor can be generalized to supporting recognition of novel categories in a novel target domain.
727 Preservation of Anomalous Subgroups On Variational Autoencoder Transformed Data S. C. Maina et al. We present a Utility Guaranteed Deep Privacy (UGDP) system which casts existing anomalous pattern detection methods as a new utility measure for data synthesis.
728 Learn-By-Calibrating: Using Calibration As A Training Objective J. J. Thiagarajan, B. Venkatesh and D. Rajan We propose a novel algorithm that performs simultaneous interval estimation for different calibration levels and effectively leverages the intervals to refine the mean estimates.
729 ESRGAN+ : Further Improving Enhanced Super-Resolution Generative Adversarial Network N. C. Rakotonirina and A. Rasoanaivo In this fashion, the model is extended to further improve the perceptual quality of the images.
730 Attentive Cutmix: An Enhanced Data Augmentation Approach for Deep Learning Based Image Classification D. Walawalkar, Z. Shen, Z. Liu and M. Savvides In this paper, we propose Attentive CutMix, a naturally enhanced augmentation strategy based on CutMix [3].
731 Efficient Image Super Resolution Via Channel Discriminative Deep Neural Network Pruning Z. Hou and S. Kung In order to identify and remove such uninformative channels, we propose a new pruning criterion, Discriminant Information, by characterizing the dependency of the output w.r.t to the hidden-layer feature-maps.
732 Multi-Resolution Overlapping Stripes Network for Person Re-Identification A. E. Okay, M. AlGhamdi, R. Westendrop and M. Abdel-Mottaleb This paper presents a part-based model with a multi-resolution network that uses different level of features.
733 Person Identification Using Deep Convolutional Neural Networks on Short-Term Signals from Wearable Sensors G. Retsinas, P. P. Filntisis, N. Efthymiou, E. Theodosis, A. Zlatintsi and P. Maragos In this work, we explore the discriminating ability of short-term signal patterns (e.g. few minutes long) with respect to the person identification task.
734 Local-Global Feature for Video-Based One-Shot Person Re-Identification C. Zhao, Z. Zhang, J. Yan and Y. Yan In this paper we propose a novel local-global progressive learning framework to overcome the limitations.
735 Global and Local Discriminative Patches Exploiting for Action Recognition J. Wu, W. Luo, W. Liu and C. Zhang In this work we propose a novel multi-stream features fusion framework based on discriminative patch exploiting.
736 Disentangling Controllable Object Through Video Prediction Improves Visual Reinforcement Learning Y. Zhong, A. Schwing and J. Peng Leveraging action-conditioned video prediction, we propose an end-to-end learning frame-work to disentangle the controllable object from the observation signal.
737 Dynamic Variational Autoencoders for Visual Process Modeling A. Sagel and H. Shen We propose a joint learning framework, combining a vector autoregressive model and a Variational Autoencoder.
738 A Novel Two-Pathway Encoder-Decoder Network for 3D Face Reconstruction X. Li, Z. Weng, J. Liang, L. C. Youjun and X. Yuli Fu To address this problem, Two-Pathway Encoder-Decoder Network (2PEDN) is proposed to regress the identity and expression components via global and local pathways.
739 Rate Assignment in 360-Degree Video Tiled Streaming Using Random Forest Regression R. Skupin, K. Bitterschulte, Y. Sanchez, C. Hellge and T. Schierl This paper addresses rate assignment in a distributed tile encoding system for such multi-resolution tiled streaming services based on the emerging Versatile Video Coding Standard.
740 Improving Convergent Cross Mapping for Causal Discovery with Gaussian Processes G. Feng, K. Yu, Y. Wang, Y. Yuan and P. M. Djuric In this paper, we propose a more reliable and principled approach, which is based on Gaussian processes (GPs) that improves the attractor reconstruction.
741 Label Reuse for Efficient Semi-Supervised Learning T. Hsieh, J. Chen and C. Chen In this paper, we propose a new learning strategy for semi-supervised deep learning algorithms, called label reuse, aiming to significantly reduce the expensive computational cost of pseudo label generation and the like for each unlabeled training instance since pseudo labels require to be repeatedly evaluated through the whole training process.
742 Decentralized Optimization with Non-Identical Sampling in Presence of Stragglers T. Adikari and S. Draper We propose to combine worker outputs weighted by the amount of work completed by each.
743 Content Vs Context: How About “Walking Hand-In-Hand” For Image Clustering? S. Hu, Z. Hou, Z. Lou and Y. Ye This paper proposes a novel content-context information bottleneck (C2IB) algorithm, which simultaneously explores and exploits the content and context information for discovering image clusters.
744 Fixed Smooth Convolutional Layer for Avoiding Checkerboard Artifacts in CNNS Y. Kinoshita and H. Kiya In this paper, we propose a fixed convolutional layer with an order of smoothness not only for avoiding checkerboard artifacts in convolutional neural networks (CNNs) but also for enhancing the performance of CNNs, where the smoothness of its filter kernel can be controlled by a parameter.
745 This Dataset Does Not Exist: Training Models from Generated Images V. Besnier, H. Jain, A. Bursuc, M. Cord and P. P?rez In this work we investigate this question and its related challenges. We identify ways to improve significantly the performance over naive training on randomly generated images with regular heuristics.
746 LEt-SNE: A Hybrid Approach to Data Embedding and Visualization Of Hyperspectral Imagery M. Shukla, B. Banerjee and K. M. Buddhiraju In this paper, we propose a novel approach, LEt-SNE, which combines graph based algorithms like t-SNE and Laplacian Eigenmaps into a model parameterized by a shallow feed forward network.
747 Adversarial Mixup Synthesis Training for Unsupervised Domain Adaptation Y. Tang, Z. Lin, H. Wang and L. Xu In this paper, we propose a new approach termed Adversarial Mixup Synthesis Training (AMST) to alleviate the issue.
748 Rate-Invariant Autoencoding of Time-Series K. Koneripalli, S. Lohit, R. Anirudh and P. Turaga In this paper, we extend prior work in disentangling latent spaces of autoencoding models, to design a novel architecture to learn rate-invariant latent codes in a completely unsupervised fashion.
749 Self-Paced Probabilistic Principal Component Analysis For Data With Outliers B. Zhao, X. Xiao, W. Zhang, B. Zhang, G. Gan and S. Xia To alleviate this problem, we propose a novel method called Self-Paced Probabilistic Principal Component Analysis (SP-PPCA) by introducing the Self-Paced Learning mechanism into PPCA.
750 Corrdrop: Correlation Based Dropout for Convolutional Neural Networks Y. Zeng, T. Dai and S. Xia To address these issues, we propose a novel structural dropout method, Correlation based Dropout (CorrDrop), to regularize CNNs by dropping feature units based on feature correlation, which reflects the discriminative information in feature maps.
751 Witchcraft: Efficient PGD Attacks with Random Step Size P. Chiang et al. We propose a variant of Projected Gradient Descent (PGD) that uses a random step size to improve performance without resorting to expensive random restarts.
752 The Fifthnet Chroma Extractor K. O. Hanlon and M. B. Sandler We propose the FifthNet, a neural network for chroma-based ACR that incorporates known spectral structures in its design through data manipulation.
753 Robust Marine Buoy Placement for Ship Detection Using Dropout K-Means Y. Ng, J. M. Pereira, D. Garagic and V. Tarokh In this paper, we formulate marine buoy placement as a clustering problem, and propose dropout k-means and dropout k-median to improve placement robustness to buoy disruption.We simulated the passage of ships in the Gabonese waters near West Africa using historical Automatic Identification System (AIS) data, then compared the ship detection probability of dropout k-means to classic k-means and dropout k-median to classic k-median.
754 On?The?Fly Feature Selection and Classification with Application to Civic Engagement Platforms Y. W. Liyanage, D. Zois and C. Chelmis Instead, we propose an optimal framework to perform joint feature selection and classification on?the?fly while relaxing the assumption on feature independence.
755 Global Traffic State Recovery VIA Local Observations with Generative Adversarial Networks M. He, X. Luo, Z. Wang, F. Yang, H. Qian and C. Hua In this paper, in order to avoid these communication overheads among spatially distributed intersections, we propose to recover the global traffic state at each intersection in a real-time fashion by only utilizing the traffic state observed at the local intersection.
756 Forecasting Sparse Traffic Congestion Patterns Using Message-Passing RNNS S. R. Iyer, U. An and L. Subramanian In this work, we leverage mobility traces of public transport vehicles tracked by the New York City MTA and formulate Message-Passing Recurrent Neural Nets (MPRNN) to produce long-term traffic forecasting on data that is sparse but wide in coverage.
757 Energy Disaggregation from Low Sampling Frequency Measurements Using Multi-Layer Zero Crossing Rate P. A. Schirmer and I. Mporas In this article we propose an extension of the zero crossing rate as a time domain measurement of frequency content and compare it to the previously proposed Karhunen Loeve Expansion, in terms of energy disaggregation performance and in terms of computational time.
758 Decoding 5G-NR Communications VIA Deep Learning P. Henarejos and M. ?ngel V?zquez In this paper we propose to use Autoencoding Neural Networks (ANN) jointly with a Deep Neural Network (DNN) to construct Autoencoding Deep Neural Networks (ADNN) for demapping and decoding.
759 Body Movement Generation for Expressive Violin Performance Applying Neural Networks J. Liu, H. Lin, Y. Huang, H. Kao and L. Su In this paper, we take a divide-and-rule approach to tackle the multifaceted characteristics of musical movement, and propose a framework for generating violinists? body movements.
760 Sequential Vessel Trajectory Identification Using Truncated Viterbi Algorithm Z. Dong, Y. Yangv and Y. Xie In this work, we propose a novel classification algorithm that used to classify vessel data points into different trajectories.
761 A Prototypical Triplet Loss for Cover Detection G. Doras and G. Peeters We thus introduce here a new test set incorporating these constraints, and propose two contributions to improve our model?s accuracy under these stricter conditions: we replace dominant melody with multi-pitch representation as input data, and describe a novel prototypical triplet loss designed to improve covers clustering.
762 Automotive Radar Signal Interference Mitigation Using RNN with Self Attention J. Mun, S. Ha and J. Lee In this paper, we propose a new method using deep learning.
763 A Large-Scale Deep Architecture for Personalized Grocery Basket Recommendations A. Mantha et al. In this paper, we introduce a production within-basket grocery recommendation system, RTT2Vec, which generates real-time personalized product recommendations to supplement the user?s current grocery basket.
764 Blind Bounded Source Separation Using Neural Networks with Local Learning Rules A. T. Erdogan and C. Pehlevan To separate such bounded sources from their mixtures, we propose a new optimization problem, Bounded Similarity Matching (BSM).
765 Modeling Piece-Wise Stationary Time Series D. Wu, S. Gundimeda, S. Mou and C. J. Quinn We propose a new, data-driven technique to automatically identify change-points and learn piece-wise stationary models.
766 Multivariate Tropical Regression and Piecewise-Linear Surface Fitting P. Maragos and E. Theodosis In this paper we propose a novel approach for multivariate convex regression by using as approximation model a maximum of hyperplanes, which we represent as a multivariate max-plus tropical polynomial.
767 Revealing Backdoors, Post-Training, in DNN Classifiers via Novel Inference on Optimized Perturbations Inducing Group Misclassification Z. Xiang, D. J. Miller and G. Kesidis We propose a defense against imperceptible backdoor attacks based on perturbation optimization and novel, robust detection inference.
768 Classifying Partially Labeled Networked Data VIA Logistic Network Lasso N. Tran, H. Ambos and A. Jung We apply the network Lasso to classify partially labeled data points which are characterized by high-dimensional feature vectors.
769 Neural Time Warping for Multiple Sequence Alignment K. Kawano, T. Kutsuna and S. Koide In this paper, we propose neural time warping (NTW) that relaxes the original MSA to a continuous optimization, in which a neural network is used to model the alignment.
770 Lance: Efficient Low-Precision Quantized Winograd Convolution for Neural Networks Based on Graphics Processing Units G. Li, L. Liu, X. Wang, X. Ma and X. Feng In this paper, we propose an efficient low-precision quan-tized Winograd convolution algorithm, called LANCE, which combines the advantages of fast convolution and quantization techniques.
771 Media Classification with Bayesian Optimization and Vapnik-Chervonenkis (VC) Bounds S. Bharitkar In this paper, we extend the approach to optimizing the class condition distribution parameters, the neural network hyperparameters, and the size of the synthetic metadata training set using a combination of Bayesian optimization and by invoking the VC-dimensionality.
772 Batman: Bayesian Target Modelling For Active Inference M. T. Koudahl and B. de Vries The main contribution of this paper is the design of a coupled generative model structure that facilitates learning desired future observations for Active Inference agents and supports integration of Active Inference and classical methods in a joint framework.
773 Deep Learning Abilities to Classify Intricate Variations in Temporal Dynamics of Multivariate Time Series P. Liotet, P. Abry, R. Leonarduzzi, M. Senneret, L. Jaffr?s and G. Perrin The aim of this work is to investigate the ability of deep learning (DL) architectures to learn temporal dynamics in multivariate time series.
774 Assimilation-Based Learning of Chaotic Dynamical Systems from Noisy and Partial Data D. Nguyen, S. Ouala, L. Drumetz and R. Fablet We propose a novel framework, which combines data assimilation schemes and neural network representation, namely Auto-Encoders and Ensemble Kalman Smoother, to learn the governing equations of dynamical systems.
775 Gated Multi-Layer Convolutional Feature Extraction Network for Robust Pedestrian Detection T. Liu, J. Huang, T. Dai, G. Ren and T. Stathaki In this paper, we propose a gated multi-layer convolutional feature extraction method which can adaptively generate discriminative features for candidate pedestrian regions.
776 Kernel Ridge Regression with Autocorrelation Prior: Optimal Model and Cross-Validation A. Tanaka and H. Imai Kernel regression problem with autocorrelation prior is discussed in this paper.
777 Generalized Kernel-Based Dynamic Mode Decomposition P. H?as, C. Herzet and B. Comb?s In this work, we devise an algorithm based on low rank constraint optimization and kernel-based computation that generalizes a recent approach called “kernel-based dynamic mode decomposition”.
778 An Online Kernel Scalar Quantization Scheme for Signal Classification J. Guo, R. G. Raj and D. J. Love In this paper, we are interested in understanding the design and behavior of these relay-like classification nodes.
779 Self-Driven Graph Volterra Models for Higher-Order Link Prediction M. Coutino, G. V. Karanikolas, G. Leus and G. B. Giannakis Cross-fertilizing ideas from Volterra series and linear structural equation models, the present paper introduces self-driven graph Volterra models that can capture higher-order interactions among nodal observables available in networked data.
780 Graph Construction from Data by Non-Negative Kernel Regression S. Shekkizhar and A. Ortega We propose non-negative kernel regression (NNK), an improved approach for graph construction with interesting geometric and theoretical properties.
781 Structured Citation Trend Prediction Using Graph Neural Networks D. Cummings and M. Nassar We present GNN-based architecture that predicts the top set of papers at the time of publication.
782 Revisiting Fast Spectral Clustering with Anchor Graph C. Wang, F. Nie, R. Wang and X. Li In this paper, we revisit the popular large-scale spectral clustering method based on the anchor graph which is equivalent to the spectral decomposition on a similar matrix obtained using a second-order transition probability.
783 A Graph Network Model for Distributed Learning with Limited Bandwidth Links and Privacy Constraints J. Parras and S. Zazo In this work, we develop an algorithm based on graph networks to train distributedly a deep learning model.
784 Graph Regularized Tensor Train Decomposition S. E. Sofuoglu and S. Aviyente In this paper, we propose a graph regularized tensor train (GRTT) decomposition that learns a low-rank tensor train model that preserves the local relationships between tensor samples.
785 Weighted Krylov-Levenberg-Marquardt Method for Canonical Polyadic Tensor Decomposition P. Tichavsk?, A. Phan and A. Cichocki The proposed Krylov-Levenberg-Marquardt method enables to do second-order-based iterations even in large-scale decomposition problems, with or without weights.
786 Low-Complexity Levenberg-Marquardt Algorithm for Tensor Canonical Polyadic Decomposition K. Huang and X. Fu In this paper, we propose CPD-fLM++, a fast implementation of the Levenberg-Marquardt (LM) algorithm for the tensor canonical polyadic decomposition.
787 A Moment-Based Approach for Guaranteed Tensor Decomposition A. Marmin, M. Castella and J. Pesquet This paper presents a new scheme to perform the canonical polyadic decomposition (CPD) of a symmetric tensor.
788 Learning Diverse Sub-Policies via a Task-Agnostic Regularization on Action Distributions L. Huo, Z. Wang, M. Xu and Y. Song In this paper, we formulate the discovery of diverse sub-policies as a trajectory inference.
789 Federated Learning with Mutually Cooperating Devices: A Consensus Approach Towards Server-Less Model Optimization S. Savazzi, M. Nicoli, V. Rampa and S. Kianoush In this paper we propose a distributed FL approach that performs a decentralized fusion of local model parameters by leveraging mutual cooperation between the devices and local (in-network) data operations via consensus-based methods.
790 No-Regret Non-Convex Online Meta-Learning Z. Zhuang, Y. Wang, K. Yu and S. Lu In this paper, we generalize the original framework from convex to non-convex setting, and introduce the local regret as the alternative performance measure.
791 Asynchrounous Decentralized Learning of a Neural Network X. Liang, A. M. Javid, M. Skoglund and S. Chatterjee In this work, we exploit an asynchronous computing framework namely ARock to learn a deep neural network called self-size estimating feedforward neural network (SSFN) in a decentralized scenario.
792 Learning Perception and Planning With Deep Active Inference O. ?atal, T. Verbelen, J. Nauta, C. D. Boom and B. Dhoedt In this paper we use recent advances in deep learning to learn the state space and approximate the necessary probability distributions to engage in active inference.
793 Projection Free Dynamic Online Learning D. S. Kalhan, A. S. Bedi, A. Koppel, K. Rajawat, A. Gupta and A. Banerjee To avoid this bottleneck, we propose a projection-free scheme based on Frank-Wolfe: where instead of online gradient steps, we use steps that are collinear with the gradient but guaranteed to be feasible.
794 Learning Partial Differential Equations From Data Using Neural Networks A. Hasan, J. M. Pereira, R. Ravier, S. Farsiu and V. Tarokh We introduce a regularization scheme that prevents the function approximation from overfitting the data and forces it to be a solution of the underlying PDE.
795 Active Learning with Unsupervised Ensembles of Classifiers P. A. Traganitis, D. Berberidis and G. B. Giannakis The present work introduces a simple scheme for active classification of data using unsupervised ensembles of classifiers.
796 Nasil: Neural Architecture Search with Imitation Learning F. S. Fard, A. Rad and V. S. Tomar Here we introduce neural architecture search with imitation learning (NASIL) method that starts the search by learning from hand designed structures by experts.
797 Multi-View Clustering Via Mixed Embedding Approximation D. Wu, F. Nie, R. Wang and X. Li Formally, we aim to learn a uniform orthogonal embedding based on the orthogonal pre-embeddings of each view.
798 Signal Clustering With Class-Independent Segmentation S. Gasperini, M. Paschali, C. Hopke, D. Wittmann and N. Navab In this paper we propose a Deep Learning-based clustering method, which encodes concurrent signals into images, and, for the first time, tackles clustering with image segmentation.
799 Mango: A Python Library for Parallel Hyperparameter Tuning S. S. Sandha, M. Aggarwal, I. Fedorov and M. Srivastava To address these challenges, we present Mango, a Python library for parallel hyperparameter tuning.
800 Anytime Minibatch with Delayed Gradients: System Performance and Convergence Analysis H. Al-Lawati and S. C. Draper We present convergence analysis of Anytime Minibatch with Delayed Gradients (AMB-DG) algorithm.
801 On Exponentially Consistency of Linkage-Based Hierarchical Clustering Algorithm Using Kolmogrov-Smirnov Distance T. Wang, Y. Liu and B. Chen This paper focuses on performance analysis of linkage-based hierarchical agglomerative clustering algorithms for sequence clustering using the Kolmogrov-Smirnov distance.
802 A Neural Network Based on First Principles P. M. Baggenstoss In this paper, a Neural network is derived from first principles, assuming only that each layer begins with a linear dimensionreducing transformation.
803 AL2: Progressive Activation Loss for Learning General Representations in Classification Neural Networks M. E. Helou, F. Dumbgen and S. S?sstrunk We propose a novel regularization method that progressively penalizes the magnitude of activations during training.
804 Label Propagation Adaptive Resonance Theory for Semi-Supervised Continuous Learning T. Kim, I. Hwang, G. Kang, W. Choi, H. Kim and B. Zhang In this paper, we propose Label Propagation Adaptive Resonance Theory (LPART) for semi-supervised continuous learning.
805 A Probabilistic Scheme for Representation Learning with Radial Transform Images H. Salehinejad and S. Valaee In this paper, a probabilistic framework to analyze radial transform is presented.
806 Perception-Distortion Trade-Off with Restricted Boltzmann Machines C. Cannella, J. Ding, M. Soltani, Y. Zhou and V. Tarokh In this work, we introduce a new procedure for applying Restricted Boltzmann Machines (RBMs) to missing data inference tasks, based on linearization of the effective energy function governing the distribution of observations.
807 Anefficient Alternative to Network Pruning Through Ensemble Learning M. P?llot, R. Zhang and A. Kaup Unlike normal pruning methods, we investigate the possibility of replacing a full-sized convolutional neural network with an ensemble of its narrow versions.
808 A Novel Pruning Approach for Bagging Ensemble Regression Based on Sparse Representation A. E. Khorashadi-Zadeh, M. Babaie-Zadeh and C. Jutten This work aims to propose an approach for pruning a bagging ensemble regression (BER) model based on sparse representation, which we call sparse representation pruning (SRP).
809 K-Autoencoders Deep Clustering Y. Opochinsky, S. E. Chazan, S. Gannot and J. Goldberger In this study we propose a deep clustering algorithm that extends the k-means algorithm.
810 MoGA: Searching Beyond Mobilenetv3 X. Chu, B. Zhang and R. Xu Bearing the target hardware in mind, we propose the first Mobile GPU-Aware (MoGA) neural architecture search in order to be precisely tailored for real-world applications.
811 Meta Metric Learning for Highly Imbalanced Aerial Scene Classification J. Guan, J. Liu, J. Sun, P. Feng, T. Shuai and W. Wang In this paper, we propose a random finetuning meta metric learning model (RF-MML) to address this problem.
812 Synthetic Crowd and Pedestrian Generator for Deep Learning Problems A. Khadka, P. Remagnino and V. Argyriou In this paper, therefore, a graphics simulator is presented which automatically generates multi-model datasets in real-time providing the corresponding ground truth and annotation.
813 TOSO: Student?s-T Distribution Aided One-Stage Orientation Target Detection in Remote Sensing Images P. Feng, Y. Lin, J. Guan, G. He, H. Shi and J. Chambers In this paper, a robust Student?s-T distribution aided One-Stage Orientation detector, namely TOSO, is proposed to address orientation target detection in remote sensing images.
814 Improving Deep Learning Classification of JPEG2000 Images Over Bandlimited Networks L. D. Chamain and Z. Ding Considering limited network bandwidth, we propose an end-to-end deep learning framework to achieve faster and more accurate classification by directly training a deep CNN image classifier using the CDF 9/7 Discrete Wavelet Transformed (DWT) coefficients from j2k-compressed images without image reconstruction.
815 Augmented Grad-CAM: Heat-Maps Super Resolution Through Augmentation P. Morbidelli, D. Carrera, B. Rossi, P. Fragneto and G. Boracchi We present Augmented Grad-CAM, a general framework to provide a high-resolution visual explanation of CNN outputs.
816 BBA-NET: A Bi-Branch Attention Network For Crowd Counting Y. Hou et al. To this end, we propose a Bi-Branch Attention Network (BBA-NET) for crowd counting, which has three innovation points.
817 Deep Metric Learning Based On Center-Ranked Loss for Gait Recognition J. Su, Y. Zhao and X. Li Therefore, in this paper, a novel loss named Center-ranked is proposed to integrate all positive and negative samples information.
818 Channel Attention Based Generative Network for Robust Visual Tracking Y. Hu, H. Xuan, J. Yang and Y. Yan In this paper, we propose a novel real-time Channel Attention based Generative Network (AGSNet) for Robust Visual Tracking.
819 Cross-VAE: Towards Disentangling Expression from Identity For Human Faces H. Wu, J. Jia, L. Xie, G. Qi, Y. Shi and Q. Tian In this paper, we propose to learn clearly disentangled and discriminative features that are invariant of identities for expression recognition.
820 Enhance Part-Based Model for Person Re-Identification with Fused Multi-Scale Features X. Lin, Y. Yang and Z. Niu To address the part-misalignment problem and learn a more discriminative embedding for person Re-ID, we propose a novel Part-based model with fused Multi-Scale features (PMS), which innovatively upscales the low-layer features by using UpShuffle Modules and smoothly integrates the high-layer features.
821 Text-To-Image Synthesis Method Evaluation Based On Visual Patterns W. L. Sommer and A. Iosifidis In this paper, we introduce an evaluation metric and a visual evaluation method allowing for the simultaneous estimation of the realism, variety and semantic accuracy of generated images.
822 Detection of Mild Dyspnea from Pairs of Speech Recordings S. Boelders, V. S. Nallanthighal, V. Menkovski and A. H?rm? In this paper, we explore techniques of detecting mild dyspnea directly from conversational speech, for example, in a telehealth application.
823 A Hybrid Model for Bipolar Disorder Classification from Visual Information N. Abaei and H. A. Osman Our goal is to classify the condition of patients suffering from BD into the clinically significant states of remission, hypo-mania, and mania.
824 Automatic Event Detection of REM Sleep Without Atonia From Polysomnography Signals Using Deep Neural Networks P. Wallis, D. Yaeger, A. Kain, X. Song and M. Lim We develop a novel, efficient, and objective method using deep learning to detect RSWA events from polysomnography signals using a large cohort of 692 patients.
825 A Deep Learning Architecture for Epileptic Seizure Classification Based on Object and Action Recognition T. Kar?csony, A. M. Loesch-Biffar, C. Vollmar, S. Noachtar and J. P. S. Cunha In this paper an end-to-end deep learning approach is proposed for binary classification of Frontal vs. Temporal Lobe Epilepsies based solely on seizure videos.
826 Transforming Seismocardiograms Into Electrocardiograms by Applying Convolutional Autoencoders M. Haescher, F. H?pfner, W. Chodan, D. Kraft, M. Aehnelt and B. Urban Heart.AI presents a fundamentally new approach, transforming motion-based seismocardiograms into electrocardiograms interpretable by cardiologists.
827 Improved Nearest Neighbor Density-Based Clustering Techniques with Application to Hyperspectral Images C. Cariou, K. Chehdi and S. Le Moan In this paper, we propose some improvements of recently published methods in this vein, namely GWENN (Graph WatershEd using Nearest Neighbors) as well as a KNN version of Density Peaks Clustering.
828 Object Surface Estimation from Radar Images O. Bialer, D. Shapiro and A. Jonas In this paper we develop a deep neural network (DNN) method for estimating the object surface from radar 2D image (azimuth-range).
829 Counting Dense Objects in Remote Sensing Images G. Gao, Q. Liu and Y. Wang In this paper, we are interested in counting dense objects from remote sensing images.
830 HPRNN: A Hierarchical Sequence Prediction Model for Long-Term Weather Radar Echo Extrapolation J. Jing, Q. Li, X. Peng, Q. Ma and S. Tang In this paper, to meet the demand for long-term extrapolation in actual forecasting practice, we propose a hierarchical prediction recurrent neural network (HPRNN) for long-term radar echo extrapolation.
831 Accurate 6D Object Pose Estimation by Pose Conditioned Mesh Reconstruction P. Castro, A. Armagan and T. Kim In order to achieve this, we propose a learning framework in which a Graph Convolutional Neural Network reconstructs a Pose Conditioned 3D mesh of the object.
832 CPWC: Contextual Point Wise Convolution for Object Recognition P. Mazumder, P. Singh and V. Namboodiri We propose an alternative design for pointwise convolution, which uses spatial information from the input efficiently.
833 Electric Analog Circuit Design with Hypernetworks And A Differential Simulator M. Rotman and L. Wolf In this work, we make a significant step towards a fully automatic design method that is based on deep learning.
834 Multi-Task Learning Via SA-FPN and EJ-Head F. Ni, Z. Luo, X. Cao, Z. Xu and Y. Yao So we propose a unified head module named EJ-Head (Effective Joint Head) to combine two branches into one head, not only realizing the interaction between two tasks, but also enhancing the effectiveness of multi-task learning.
835 Differentiable Branching In Deep Networks for Fast Inference S. Scardapane, D. Comminiello, M. Scarpiniti, E. Baccarelli and A. Uncini In this paper, we consider the design of deep neural networks augmented with multiple auxiliary classifiers departing from the main (backbone) network.
836 Multi-Step Online Unsupervised Domain Adaptation J. H. Moon, D. Das and C. S. G. Lee In this paper, we address the Online Unsupervised Domain Adaptation (OUDA) problem, where the target data are unlabelled and arriving sequentially.
837 Self-Adaptive Feature Fool X. Liu, Y. Bai, S. Xia and Y. Jiang In this paper, we propose a novel method to craft UAPs in the absence of data, via adaptively perturbing mid-layer outputs of the CNN.
838 Multi-MotifGAN (MMGAN): Motif-Targeted Graph Generation And Prediction A. Gamage, E. Chien, J. Peng and O. Milenkovic We propose Multi-MotifGAN (MMGAN), a motif-targeted Generative Adversarial Network (GAN) that generalizes the benchmark NetGAN approach.
839 Federated Classification with Low Complexity Reproducing Kernel Hilbert Space Representations M. Peifer and A. Ribeiro In this work we propose a method for federated learning in which each agent learns a low complexity Reproducing kernel Hilbert space representation.
840 Maxpolynomial Division with Application To Neural Network Simplification G. Smyrnis, P. Maragos and G. Retsinas To that end, we introduce the process of Maxpolynomial Division, a geometric method which simulates division of polynomials in the max-plus semiring, while highlighting its key properties and noting its connection to neural networks.
841 Balanced Binary Neural Networks with Gated Residual M. Shen, X. Liu, R. Gong and K. Han In this paper, we attempt to maintain the information propagated in the forward process and propose a Balanced Binary Neural Networks with Gated Residual (BBG for short).
842 A Geometric Approach for Unsupervised Similarity Learning U. K. Dutta and C. Sekhar C In this paper, we propose a novel, unsupervised metric learning approach, that learns a similarity metric without making use of class labels.
843 Gradient Delay Analysis in Asynchronous Distributed Optimization H. Al-Lawati and S. C. Draper In this work, we focus on the asynchronous implementation of gradient-based algorithms.
844 Sequential IoT Data Augmentation Using Generative Adversarial Networks M. E. Tschuchnig, C. Ferner and S. Wegenkittl This paper investigates the possibility of using GANs in order to augment sequential Internet of Things (IoT) data, with an example implementation that generates household energy consumption data with and without swimming pools.
845 Robust Rank Constrained Sparse Learning: A Graph-Based Method for Clustering R. Liu, M. Chen, Q. Wang and X. Li To solve this problem, a robust rank constrained sparse learning method is proposed in this paper.
846 Efficient Decoupled Neural Architecture Search by Structure And Operation Sampling H. Lee, D. Kim and B. Han We propose a novel neural architecture search algorithm via reinforcement learning by decoupling structure and operation search.
847 Weight Sharing and Deep Learning for Spectral Data J. S. Larsen and L. Clemmensen We propose a novel method to co-train deep convolutional neural networks for data sets of differing position specific data.
848 Complex Transformer: A Framework for Modeling Complex-Valued Sequence M. Yang, M. Q. Ma, D. Li, Y. H. Tsai and R. Salakhutdinov In this paper, we propose a Complex Transformer, which incorporates the transformer model as a backbone for sequence modeling; we also develop attention and encoder-decoder network operating for complex input.
849 High-Dimensional Neural Feature Using Rectified Linear Unit And Random Matrix Instance A. M. Javid, A. Venkitaraman, M. Skoglund and S. Chatterjee We design a ReLU-based multilayer neural network to generate a rich high-dimensional feature vector.
850 Projected Weight Regularization to Improve Neural Network Generalization G. Zhang, K. Niwa and W. B. Kleijn In this paper we propose a new technique, named projected weight regularization (PWR), to improve the generalization capacity of a DNN model.
851 Deep Clustering for Domain Adaptation B. Gao, Y. Yang, H. Gouk and T. M. Hospedales We address the heterogeneous domain adaptation task: adapting a classifier trained on data from one domain to operate on another domain that also has a different label space.
852 Deep Clusteringwith Concrete K-Means B. Gao, Y. Yang, H. Gouk and T. M. Hospedales We address the problem of simultaneously learning a k-means clustering and deep feature representation from unlabelled data, which is of interest due to the potential for deep k-means to outperform traditional two-step feature extraction and shallow clustering strategies.
853 Polarizing Front Ends for Robust Cnns C. Bakiskan, S. Gopalakrishnan, M. Cekic, U. Madhow and R. Pedarsani In this paper, we propose a bottom-up strategy for attenuating adversarial perturbations using a nonlinear front end which polarizes and quantizes the data.
854 Adaptive Distributed Stochastic Gradient Descent for Minimizing Delay in the Presence of Stragglers S. K. Hanna, R. Bitar, P. Parag, V. Dasari and S. El Rouayheb Towards optimizing the error-runtime trade-off, we investigate distributed SGD with adaptive k.
855 A Model of Double Descent for High-Dimensional Logistic Regression Z. Deng, A. Kammoun and C. Thrampoulidis We consider a model for logistic regression where only a subset of features of size p is used for training a linear classifier over n training samples.
856 Efficient Scene Text Detection with Textual Attention Tower L. Zhang et al. In this work, we propose an efficient and accurate approach to detect multi-oriented text in scene images.
857 A Hybrid Approach for Thermographic Imaging With Deep Learning P. Kov?cs, B. Lehner, G. Thummerer, G. Mayr, P. Burgholzer and M. Huemer We propose a hybrid method for reconstructing thermographic images by combining the recently developed virtual wave concept with deep neural networks.
858 Knowledge Enhanced Latent Relevance Mining for Question Answering D. Wang, Y. Shen and H. Zheng In this paper, we propose to integrate commonsense from the external knowledge graphs (KGs) into the answer selector through a knowledge-aware context-based attention mechanism.
859 Multi-Label Consistent Convolutional Transform Learning: Application to Non-Intrusive Load Monitoring S. Singh, J. Maggu, A. Majumdar, E. Chouzenoux and G. Chierchia In this work, we propose a supervised formulation for convolutional transform so as to address the multi-label classification problem.
860 Resilient Distributed Recovery of Large Fields Y. Chen, S. Kar and J. M. F. Moura We present a field recovery consensus+innovations type distributed algorithm that is resilient to measurement attacks, where an agent maintains and updates a local state based on its neighbors states and its own measurement.
861 Training LSTM for Unsupervised Anomaly Detection Without A Priori Knowledge Y. Cherdo, P. d. Kerret and R. Pawlak In contrast, we propose a novel anomaly detector, coined as LSTM-Decomposed (LSTM-D), that does not require this normality knowledge.
862 Unsupervised Person Re-Identification Using Multi-Branch Feature Compensation Network and Link-Based Cluster Dissimilarity Metric L. Pan, G. Qi, B. Guo and Y. Zhu Therefore in this paper, a Multi-branch Feature Compensation Network (MFC-Net) is developed in which the significant parts of lower-layer features are learned and fused with high-layer feature as compensation.
863 Deep-SST-Eddies: A Deep Learning Framework to Detect Oceanic Eddies in Sea Surface Temperature Images E. Moschos, O. Schwander, A. Stegner and P. Gallinari We introduce a novel method that employs Deep Learning to detect eddy signatures on such input.
864 Interpretability-Guided Convolutional Neural Networks for Seismic Fault Segmentation Z. Liu, C. Zhou, G. Hu and C. Song To include domain knowledge to improve the interpretability of the CNN, we propose to jointly optimize the prediction accuracy and consistency between explanations of the neural network and domain knowledge.
865 Towards High-Performance Object Detection: Task-Specific Design Considering Classification and Localization Separation J. U. Kim, S. T. Kim, E. S. Kim, S. Moon and Y. M. Ro In this work, we simply modified layers of the existing object detection networks into three parts by considering such characteristics: lower-layer feature sharing part, layer separation part, and feature fusion part.
866 Anomaly Detection for Time Series Using VAE-LSTM Hybrid Model S. Lin, R. Clark, R. Birke, S. Sch?nborn, N. Trigoni and S. Roberts In this work, we propose a VAE-LSTM hybrid model as an unsupervised approach for anomaly detection in time series.
867 Hydranet: A Real-Time Waveform Separation Network E. T. Kaspersen, T. Kounalakis and C. Erkut In this paper, we propose a 1-D convolutional U-Net structure to separate waveform input.
868 Storing Digital Data Into DNA: A Comparative Study Of Quaternary Code Construction M. Dimopoulou, M. Antonini, P. Barbry and R. Appuswamy In this paper we present a comparative study of our work with the state of the art solutions, and show that our solution is competitive.
869 A New Multihypothesis Prediction Scheme for Compressed Video Sensing Reconstruction S. Zheng, X. Zhang, J. Chen and Y. Kuo To solve this problem, this paper proposes a new multihypothesis prediction scheme.
870 Bit Allocation for Multi-Task Collaborative Intelligence S. R. Alvar and I. V. Bajic In this paper, we propose the first bit allocation method for multi-stream, multi-task CI.
871 Motion Feedback Design for Video Frame Interpolation M. Hu, L. Liao, J. Xiao, L. Gu and S. Satoh This paper introduces a feedback-based approach to interpolate video frames involving small and fast-moving objects.
872 Trilingual Semantic Embeddings of Visually Grounded Speech with Self-Attention Mechanisms Y. Ohishi, A. Kimura, T. Kawanishi, K. Kashino, D. Harwath and J. Glass We propose a trilingual semantic embedding model that associates visual objects in images with segments of speech signals corresponding to spoken words in an unsupervised manner.
873 Towards Pose-Invariant Lip-Reading S. Cheng et al. Specifically, we propose to use a 3D Morphable Model (3DMM) to augment LRW, an existing large-scale but mostly frontal dataset, by generating synthetic facial data in arbitrary poses.
874 A Siamese Content-Attentive Graph Convolutional Network for Personality Recognition Using Physiology H. Yang and C. Lee In this work, we propose a novel Siamese Content-Attentive Graph Convolutional Network (SCA-GCN) to learn a discriminative physiology representation jointly guided by the actual video content of the emotional stimuli.
875 Self-Supervised Learning for Audio-Visual Speaker Diarization Y. Ding, Y. Xu, S. Zhang, Y. Cong and L. Wang In this paper, we propose a self-supervised audio-video synchronization learning method to address the problem of speaker diarization without massive labeling effort.
876 What Makes the Sound?: A Dual-Modality Interacting Network for Audio-Visual Event Localization J. Ramaswamy In this paper, we address the problem of audio-visual event localization where the goal is to identify the presence of an event that is both audible and visible in a video, using fully or weakly supervised learning.
877 Attentional Fused Temporal Transformation Network for Video Action Recognition K. Yang et al. Focusing on discriminate spatiotemporal feature learning, we propose Attentional Fused Temporal Transformation Network (AttnTTN) for action recognition on top of popular Temporal Segment Network (TSN) framework.
878 Deep Product Quantization Module for Efficient Image Retrieval M. Liu, Y. Dai, Y. Bai and L. Duan We propose a simple but effective deep Product Quantization Module (PQM) to jointly learn discriminative codebook and precise hard assignment in an end-to-end manner.
879 The Open Brands Dataset: Unified Brand Detection and Recognition at Scale X. Jin, W. Su, R. Zhang, Y. He and H. Xue In order to tackle these problems, we first define the special issues of brand detection and recognition compared with generic object detection. Second, a novel brands benchmark called “Open Brands” is established.
880 Spectrogram Analysis Via Self-Attention for Realizing Cross-Model Visual-Audio Generation H. Tan, G. Wu, P. Zhao and Y. Chen In this paper, the Self-Attention mechanism is applied to cross-modal visual-audio generation for the first time.
881 DGAN: Disentangled Representation Learning for Anisotropic BRDF Reconstruction Z. Hu, X. Wang and Q. Wang In this paper, we present a novel deep architecture, Disentangled Generative Adversarial Network (DGAN), which performs anisotropic Bidirectional Reflectance Distribution Function (BRDF) reconstruction from single BRDF subspace with the maximum entropy.
882 APB2FACE: Audio-Guided Face Reenactment with Auxiliary Pose and Blink Signals J. Zhang, L. Liu, Z. Xue and Y. Liu To solve those problems, we propose a novel deep neural network named APB2Face, which consists of GeometryPredictor and FaceReenactor modules.
883 Motion Dynamics Improve Speaker-Independent Lipreading M. Riva, M. Wand and J. Schmidhuber We present a novel lipreading system that improves on the task of speaker-independent word recognition by decoupling motion and content dynamics.
884 Multi-Layer Content Interaction Through Quaternion Product for Visual Question Answering L. Shi et al. To solve the issue for the intermediate layers, we propose an efficient Quaternion Block Network (QBN) to learn interaction not only for the last layer but also for all intermediate layers simultaneously.
885 Linear Model-Based Intra Prediction in VVC Test Model R. Ghaznavi-Youvalari This paper studies a new intra prediction method based on a linear model for improving the intra prediction performance of Versatile Video Coding (H.266/VVC) standard.
886 Intra Frame Rate Control for Versatile Video Coding with Quadratic Rate-Distortion Modelling Y. Chen, S. Kwong, M. Zhou, S. Wang, G. Zhu and Y. Wang This paper proposes a new quadratic R-D model for Versatile Video Coding.
887 Performance Comparison of Lossless Compression Strategies for Dynamic Vision Sensor Data K. Iqbal, N. Khan and M. G. Martini We compare the performance of a number of strategies, including the only strategy developed specifically for such data and other more generic data compression strategies, tailored here to the case of neuromorphic data.
888 Encoder-Recurrent Decoder Network for Single Image Dehazing A. Dang, T. H. Vu and J. Wang This paper develops a deep learning model, called Encoder-Recurrent Decoder Network (ERDN), which recovers the clear image from a degrade hazy image without using the atmospheric scattering model.
889 Denoising of Event-Based Sensors with Spatial-Temporal Correlation J. Wu, C. Ma, X. Yu and G. Shi In this paper, we introduce a novel event stream denoising method for such sensors.
890 A Visual-Pilot Deep Fusion for Target Speech Separation in Multitalker Noisy Environment Y. Li, Z. Liu, Y. Na, Z. Wang, B. Tian and Q. Fu In this paper, we deploy face tracking and propose the low-dimension hand-crafted visual features and the low-cost deep fusion architectures to separate the unseen but visible target sources in multi-talker noisy environment.
891 C3DVQA: Full-Reference Video Quality Assessment with 3D Convolutional Neural Network M. Xu, J. Chen, H. Wang, S. Liu, G. Li and Z. Bai In this paper, we present a novel architecture, namely C3DVQA, that uses Convolutional Neural Network with 3D kernels (C3D) for full-reference VQA task.
892 Exploring Entity-Level Spatial Relationships for Image-Text Matching Y. Xia, L. Huang, W. Wang, X. Wei and J. Chen To this end, we utilize the relative position of objects to capture entity-level spatial relationships for image-text matching.
893 A Deep Multimodal Approach for Map Image Classification T. Sawada and M. Katsurai Specifically, we present a novel strategy for preprocessing text data that are positioned inside the map images, which are extracted using OCR.
894 Selective Convolutional Network: An Efficient Object Detector with Ignoring Background H. Ling, Y. Qin, L. Zhang, Y. Shi and P. Li Therefore, we introduce an efficient object detector called Selective Convolutional Network (SCN), which selectively calculates only on the locations that contain meaningful and conducive information.
895 Back-And-Forth Prediction for Deep Tensor Compression H. Choi, R. A. Cohen and I. V. Bajic In this paper we present a prediction scheme called Back-and-Forth (BaF) prediction, developed for deep feature tensors, which allows us to dramatically reduce tensor size and improve its compressibility.
896 Effective Pipeline for Compressing Deep Object Detectors Y. Yao, Z. Fang, B. Dong and S. Zhou To alleviate the deployment of deep object detectors with large model capacity and complex computation, an effective model compression pipeline is designed in this paper.
897 Gated Mechanism for Attention Based Multi Modal Sentiment Analysis A. Kumar and J. Vepa In this paper, we address three aspects of multimodal sentiment analysis; 1. Cross modal interaction learning, i.e. how multiple modalities contribute to the sentiment, 2. Learning long-term dependencies in multimodal interactions and 3. Fusion of unimodal and cross modal cues.
898 Multitask Learning and Multistage Fusion for Dimensional Audiovisual Emotion Recognition B. T. Atmaja and M. Akagi This paper proposes two methods to predict emotional attributes from audio and visual data using a multitask learning and a fusion strategy.
899 Object Detection and 3d Estimation Via an FMCW Radar Using a Fully Convolutional Network G. Zhang, H. Li and F. Wenger To ensure successful training of a fully convolutional network (FCN), we propose a normalization method, which is found to be essential to be applied to the radar signal before feeding into the neural network.
900 Ava Active Speaker: An Audio-Visual Dataset for Active Speaker Detection J. Roth et al. In this paper, we present the AVA Active Speaker detection dataset (AVA-ActiveSpeaker) which has been publicly released to facilitate algorithm development and comparison.
901 Supervised Deep Hashing for Efficient Audio Event Retrieval A. Jati and D. Emmanouilidou This work investigates the potency of different hashing techniques for efficient audio event retrieval.
902 An LSTM-Based Dynamic Chord Progression Generation System for Interactive Music Performance C. Garoufis, A. Zlatintsi and P. Maragos In this paper, we describe an interactive generative music system, designed to handle polyphonic guitar music.
903 Ensemble Network For Ranking Images Based On Visual Appeal S. Singh, V. Sanchez and T. Guha We propose a computational framework for ranking images (group photos in particular) taken at the same event within a short time span.
904 Trapezoidal Segment Sequencing: A Novel Approach for Fusion of Human-Produced Continuous Annotations B. M. Booth and S. S. Narayanan This work proposes trapezoidal segment sequencing, a new method for fusing annotations into a single representation that, when used alongside a recently proposed signal warping pipeline for correcting annotation artifacts, produces accurate ground truths.
905 Sequence-to-Sequence Labanotation Generation Based on Motion Capture Data M. Li, Z. Miao and C. Ma In this paper, we propose a sequence-to-sequence approach that can generate Labanotation scores from unsegmented motion data sequences.
906 Pose Refinement: Bridging the Gap Between Unsupervised Learning and Geometric Methods for Visual Odometry L. Zhang, G. Li and T. H. Li We propose a hybrid VO system which combines an unsupervised monocular VO with a pose refinement backend, performing optimization in real time, leading to boosted performance.
907 Multimodal Active Speaker Detection and Virtual Cinematography for Video Conferencing R. Cutler et al. We describe a new automated ASD and VC that performs within 0.3 MOS of an expert cinematographer based on subjective ratings with a 1-5 scale.
908 A New Variational Method for Deep Supervised Semantic Image Hashing F. Zhuang and P. Moulin We present a supervised semantic hashing method which uses a variational autoencoder to represent each database image sample as a product Bernoulli distribution.
909 DOA Estimation in Systems with Nonlinearities for MMWAVE Communications A. Sant and B. D. Rao We address the DOA estimation problem for a general nonlinear distortion of the received signal.
910 Wideband Direction of Arrival Estimation with Sparse Linear Arrays F. Wang, Z. Tian, J. Fang and G. Leus This paper concerns wideband direction of arrival (DoA) estimation with sparse linear arrays (SLAs).
911 Fourth Order Cumulant Based Active Direction of Arrival Estimation Using Coprime Arrays Z. Fu, P. Charg? and Y. Wang In this paper, we investigate the fourth order cumulant for active DOA estimation.
912 On Regularization Parameter for L0-Sparse Covariance Fitting Based DOA Estimation A. Delmer, A. Ferr?ol and P. Larzabal In this paper, we provide a statistical method allowing to estimate an admissible interval where ? must be chosen.
913 Effective Approximate Maximum Likelihood Estimation of Angles of Arrival for Non-Coherent Sub-Arrays T. Tirer and O. Bialer In this paper, we propose a technique to estimate the sub-arrays phase offsets for a given AOAs hypothesis, which facilitates approximate maximum likelihood estimation of the AOAs from a single snapshot.
914 Two-dimensional DOA Estimation for Coprime Planar Array: A Coarray Tensor-based Solution H. Zheng, C. Zhou, Y. Gu and Z. Shi To address this problem, we propose a novel coarray tensor-based two-dimensional underdetermined DOA estimation method for coprime planar array in this paper, where both the multi-dimensional information of the received signals and the augmented coarray are effectively utilized.
915 Multi-constraint Spectral Co-design for Colocated MIMO Radar and MIMO Communications S. H. Dokhanchi, M. R. Bhavani Shankar, K. V. Mishra and B. Ottersten The paper addresses the challenge of designing a waveform in MIMO-radar MIMO-communications (MRMC) set-up in a broadcast environment to ensure certain performance of the two systems is guaranteed.
916 Tensor Decomposition-based Beamspace Esprit Algorithm for Multidimensional Harmonic Retrieval F. Wen, H. C. So and H. Wymeersch In this paper, we propose a beamspace tensor-ESPRIT for multidimensional HR.
917 Information Theoretic Approach for Waveform Design in Coexisting MIMO Radar and MIMO Communications M. Alaee-Kerharoodi, S. M. R. Bhavani, K. V. Mishra and B. Ottersten We investigate waveform design for coexistence between a multiple-input multiple-output (MIMO) radar and MIMO communications (MRMC), with a radar-centric criterion that leads to a minimal interference in the communications system.
918 Transmit Beampattern Shaping via Waveform Design in Cognitive Mimo Radar E. Raei, M. Alaee-Kerahroodi, B. S. M. R. and B. Ottersten This paper is focused on designing a set of constant modulus waveform for cognitive Multiple-Input Multiple-Output (MIMO) radar systems.
919 Multilinear Generalized Singular Value Decomposition (Ml-gsvd) with Application to Coordinated Beamforming in Multi-user Mimo Systems L. Khamidullina, A. L. F. de Almeida and M. Haardt In this paper, we propose a new Multilinear Generalized Singular Value Decomposition (ML-GSVD) which allows to jointly factorize a set of matrices with one common dimension.
920 Sparse Low-redundancy Linear Array with Uniform Sum Co-array R. Rajam?ki and V. Koivunen In this paper, we introduce a parametric sparse linear array configuration called the Kl?ve array (KA).
921 Compressed Sensing Based Channel Estimation and Open-loop Training Design for Hybrid Analog-digital Massive MIMO Systems K. Ardah, B. Sokal, A. L. F. de Almeida and M. Haardt In this paper, we propose an openloop hybrid analog-digital beam-training framework, where a given sensing matrix is decomposed into analog and digital beamformers.
922 Dispersive Grid-free Orthogonal Matching Pursuit for Modal Estimation in Ocean Acoustics T. Paviet-Salomon, C. Dorffer, J. Bonnel, B. Nicolas, T. Chonavel and A. Dr?meau In the continuation of previous works, we propose here a new grid-free algorithm allowing a super-resolution of the (f-k) diagram by benifiting from the sparse nature of the wavenumber spectrum and embedding the broadband behavior of the wavenumbers within the algorithm.
923 Greedy Sparse Array Design for Optimal Localization under Spatially Prioritized Source Distribution Y. Gershon, Y. Buchris and I. Cohen We introduce a new measure called power spread, which quantifies the localization error region. Then, we propose a greedy algorithm for a sparse array design, aiming to minimize the power spread for optimal error region in a given area of interest.
924 Compressive 2-d Off-grid DOA Estimation for Propeller Cavitation Localization Y. Park and P. Gerstoft This paper introduces compressive sensing (CS) based two-dimensional (2-D) off-grid direction-of-arrival (DOA) estimation approach which can output the azimuths and elevations of radiating sources for propeller tip vortex cavitation localization.
925 The Compressed Nested Array for Underdetermined DOA Estimation by Fourth-order Difference Coarrays Y. Zhou, Y. Li, L. Wang, C. Wen and W. Nie In this paper, a new sparse array structure, which further improves the degrees of freedom (DOFs) and enhanced the DOA estimation performance, for the fourth-order cumulant based direction of arrival (DOA) estimation is proposed.
926 Model Order Selection in DoA Scenarios via Cross-entropy Based Machine Learning Techniques A. Barthelme, R. Wiesmayr and W. Utschick In this paper, we present a machine learning approach for estimating the number of incident wavefronts in a direction of arrival scenario.
927 Unsupervised Change Detection for Multimodal Remote Sensing Images via Coupled Dictionary Learning and Sparse Coding V. Ferraris, N. Dobigeon, Y. Cavalcanti, T. Oberlin and M. Chabert This paper proposes a coupled dictionary learning strategy to detect changes between two images with different modalities and possibly different spatial and/or spectral resolutions.
928 Fast Direction-of-arrival Estimation of Multiple Targets Using Deep Learning and Sparse Arrays G. K. Papageorgiou and M. Sellathurai In this work, we focus on improving the Direction-of-Arrival (DoA) estimation of multiple targets/sources from a small number of snapshots.
929 Learning Based Reconfigurable Sub-nyquist Sampling Framework for Ultra-wideband Angular Sensing H. Joshi, M. Alaee-Kerahroodi, A. A. Kumar, B. Shankar Mysore R and S. J. Darak In this work, an intelligent and reconfigurable ultra-wideband angular sensing (UWAS) framework is proposed which is independent of the maximum number of active transmissions in a wideband spectrum unlike the existing UWAS methods.
930 Raw Waveform Based End-to-end Deep Convolutional Network for Spatial Localization of Multiple Acoustic Sources H. Sundar, W. Wang, M. Sun and C. Wang In this paper, we present an end-to-end deep convolutional neural network operating on multi-channel raw audio data to localize multiple simultaneously active acoustic sources in space.
931 DNN-based Mask Estimation Integrating Spectral and Spatial Features for Robust Beamforming C. Deng, H. Song, Y. Zhang, Y. Sha and X. Li In this paper, we propose a novel spectral-spatial mask based beamforming method for two-channel noisy signals, where spectral amplitude and cross-channel spatial features are integrated to improve mask estimation.
932 Ray Separation and Source Depth Estimation Based on Sound Pressure Field Transformation R. Wei, X. Ma and X. Li In this paper, we address the issue of submerged sound source depth estimation in the deep ocean using the idea of multipath ray separation.
933 Regularized Beamformer for the Spherical Microphone Array to Cope with the White Noise Amplification L. Wang and J. Zhu This paper presents two regularization methods to optimize the performance of the high directivity beamformer for the spherical microphone array.
934 Interpolation and Range Extrapolation of Sound Source Directivity Based on a Spherical Wave Propagation Model J. Ahrens and S. Bilbao We revisit the previously proposed approach of only using the angle-dependent magnitude of the measured directivity together with a spherical-wave propagation model and demonstrate its potential by means of numerical simulations based on two case studies.
935 Polarization Parameters Estimation with Scalar Sensor Arrays M. Dai, X. Ma, W. Liu and W. Sheng The polarization sensitivity model of an SSA is first established and then as an example, a dimension-reduction method based on multiple signal classification (MUSIC) is employed to jointly estimate the direction-of-arrival and polarization parameters.
936 DNN-based Distributed Multichannel Mask Estimation for Speech Enhancement in Microphone Arrays N. Furnon, R. Serizel, I. Illina and S. Essid In this context, we propose to extend the distributed adaptive node-specific signal estimation approach to a neural network framework.
937 Asymptotically Optimal Blind Calibration of Acoustic Vector Sensor Uniform Linear Arrays A. Weiss, B. Nadler and A. Yeredor We study the blind calibration problem of uniform linear arrays of acoustic vector sensors for narrowband Gaussian signals, and propose an improved, asymptotically optimal blind calibration scheme.
938 Optimized Sensor Selection for Joint Radar-communication Systems A. Ahmed, S. Zhang and Y. D. Zhang In this paper, we address optimal sensor selection for adaptive beamforming-based JRC systems by exploiting a constrained re-weighted l1-norm minimization with low computational complexity.
939 Frequency Diverse Array Radar: A Closed-Form Solution to Design Weights for Desired Beampattern M. Zubair, S. Ahmed and M. Alouini In this work, a novel algorithm with low complexity is proposed to focus the transmitted power of an FDA radar in the desired region-of-interest for longer dwell time.
940 Robust CFAR Radar Detection Using a K-nearest Neighbors Rule A. Coluccia, A. Fascista and G. Ricci In particular, a novel interpretation of the well-known Kelly�s and adaptive matched filter (AMF) detectors is provided in terms of decision region boundaries in a suitable feature space. Then, a new detector based on a feature vector that combines the two detection statistics is obtained by exploiting the k-nearest neighbors (KNN) approach.
941 Cramer-Rao Bound on DOA Estimation of Finite Bandwidth Signals Using a Moving Sensor A. Arora, B. S. Mysore R and B. Ottersten In this paper, we provide a framework for the direction of arrival (DOA) estimation using a single moving sensor and evaluate performance bounds on estimation.
942 Theoretical Analysis of Multi-Carrier Agile Phased Array Radar T. Huang, N. Shlezinger, X. Xu, D. Ma, Y. Liu and Y. C. Eldar In this paper, we theoretically analyze the range-Doppler recovery capabilities of CAESAR.
943 Convolutional Beamspace for Array Signal Processing P. P. Vaidyanathan and P. Chen A new type of beamspace for array processing is introduced called convolutional beamspace.
944 Efficient Estimation of Mixing Matrix Using a Two-sensor Array Q. Yan, S. Sun, H. Zhang and G. Hua Considering the computational cost in practice, this paper proposes an efficient mixing matrix estimation (MME) method using an easily configured two-sensor array.
945 Riemannian Geometry and Cram?r-rao Bound for Blind Separation of Gaussian Sources F. Bouchard, A. Breloy, A. Renaux and G. Ginolhac We consider the optimal performance of blind separation of Gaussian sources.
946 Beamformed Feature for Learning-based Dual-channel Speech Separation H. Li, X. Zhang and G. Gao This paper deals with the problem of separating target speech signal from reverberant and noisy environment with dual microphones, where the target speech comes from a predefined direction range.
947 Transmit Beamforming Design with Received-Interference Power Constraints: The Zero-Forcing Relaxation E. Lagunas, A. P?rez-Neira, M. A. Lagunas and M. A. Vazquez In this paper, we study in detail an alternative transmit beamforming design framework, where we allow some residual received-interference power instead of trying to null it completely out.
948 Foreground Signature Extraction for an Intimate Mixing Model in Hyperspectral Image Classification J. Hollis, R. Raich, J. Kim, B. Fishbain and S. Kendler We introduce a framework for foreground signature extraction based on a proposed patch model.
949 Pixel-Wise Linear/Nonlinear Nonnegative Matrix Factorization for Unmixing of Hyperspectral Data F. Zhu, P. Honeine and J. Chen In this work, we propose an unsupervised nonlinear unmixing method that overcomes these weaknesses.
950 Beam Elimination Based on Sequentially Estimated a Posteriori Probabilities of Winning M. K. Marandi, W. Rave and G. Fettweis In this paper, we refine the elimination process by introducing a new elimination mechanism based on estimated winning probability i.e. probability of being the strongest candidate for each beam at each time step.
951 Two-Element Biomimetic Antenna Array Design and Performance R. J. Kozick, F. T. Dagefu and B. M. Sadler The optimum coupling network that minimizes the Cramer-Rao bound (CRB) for angle of arrival (AOA) estimation is derived and a design method is presented to synthesize the network.
952 Distributed Equalization and Power Allocation For Multi-Carrier Bidirectional Filter-and-Forward Relay Networks S. KianiHarchegani, S. ShahbazPanahi, M. Dong and G. Boudreau We propose a novel semi-closed form sub-optimal solution to this non-convex joint optimization problem.
953 Upscaling Vector Approximate Message Passing N. Skuratovs and M. Davies In this paper we consider the problem of recovering a signal x of size N from noisy and compressed measurements y = Ax + w of size M, where the measurement matrix A is right-orthogonally invariant (ROI).
954 Atomic Norm Based Localization of Far-Field and Near-Field Signals with Generalized Symmetric Arrays X. Wu, W. Zhu and J. Yan In this paper, we propose a localization method for mixed FF and NF sources based on the generalized symmetric linear arrays, which include ULAs, Cantor array, Fractal array and many other SLAs.
955 A Novel Moving Sparse Array Geometry with Increased Degrees of Freedom S. Li and X. Zhang In this paper, we propose a novel moving sparse array geometry named dilated arrays (DAs) by extending the dilation of nested arrays to other linear array structures.
956 Clutter Identification Based on Sparse Recovery and L1-Type Probabilistic Distance Measures Y. Zhu, Y. Xiang, S. Sen, E. Dagois, A. Nehorai and M. Akcakaya We form a dictionary of various clutter distributions and identify the distribution of the new clutter data through matching pursuit using probabilistic similarity and distance measures under sparsity constraints.
957 Adaptive Subspace Detectors for off-grid Mismatched Targets J. Bosse, O. Rabaste and J. Ovarlez We propose here to use adaptive subspace detectors to solve this issue, a suitable subspace (that coincides with the Discrete Prolate Spheroidal Sequences basis when the signal model is that of sinusoids in noise) is proposed that offers robust performance.
958 A Method for Millimeter-Wave Imaging of Concealed Objects Via De-Aliasing W. Wang and K. Yang In this paper, we propose a new imaging method by de-aliasing.
959 A Fast Sparse Covariance-Based Fitting Method for DOA Estimation via Non-Negative Least Squares C. Zheng, H. Chen and A. Wang A fast sparse covariance-based fitting algorithm with the non-negative least squares (NNLS) form is proposed for the direction of arrival (DOA) estimation.
960 Extended Cyclic Coordinate Descent for Robust Row-Sparse Signal Reconstruction in the Presence of Outliers H. Huang, H. C. So and A. M. Zoubir We propose an extended CCD algorithm to solve the problem for complex-valued measurements, which requires careful characterization and derivation.
961 Low-Rank Toeplitz Matrix Estimation Via Random Ultra-Sparse Rulers H. Lawrence, J. Li, C. Musco and C. Musco In this work, we introduce random ultra-sparse rulers and propose an improved approach based on these objects.
962 Urtis: a Small 3d Imaging Sonar Sensor for Robotic Applications T. Verellen, R. Kerstens, D. Laurijssen and J. Steckel In this paper we will discuss the newest 3D in-air sonar sensor developed by CoSys-Lab: the micro Real Time Imaging Sonar (?RTIS).
963 A Partial Relaxation DOA Estimator Based on Orthogonal Matching Pursuit M. Trinh-Hoang, W. Ma and M. Pesavento By adopting the principle of the partial relaxation approach, in this paper, a modification of the Orthogonal Matching Pursuit algorithm is proposed and applied to the Direction-of-Arrival estimation problem.
964 Variable Projection for Multiple Frequency Estimation Y. E. Garcia, P. Kov?cs and M. Huemer In this paper, such formulation is derived, and a variable projection (VP) optimization is proposed for solving the SNLLS problem and estimate the frequency parameters.
965 Fusionndvi: A Novel Fusion Method for NDVI in Remote Sensing M. Zhang, Z. Zhao, Y. Chen, Z. Wang and X. Tian The fusion problem is formulated to minimize a least square fitting error term and a nonlocal gradient sparsity regularization term.
966 Robust Music Estimation Under Array Response Uncertainty A. Bazzi We introduce a MUSIC estimator, hereby referred to as Robust-MUSIC, that is capable of estimating AoAs of multiple signals, when the antenna array?s response is subject to uncertainties, due to the aforementioned reasons.
967 L1-Norm Higher-Order Orthogonal Iterations for Robust Tensor Analysis D. G. Chachlakis, A. Prater-Bennette and P. P. Markopoulos To counteract the impact of outliers in tensor data analysis, we propose L1-Tucker: a reformulation of standard Tucker decomposition, resulting by simple substitution of the outlier-responsive L2-norm by the sturdier L1-norm.
968 Sensor Selection for Model-Free Source Localization: where Less is More E. Tohidi, J. Chen and D. Gesbert In this paper, we demonstrate that instead of employing all the sensors that result in a possibly unbalanced sensing pattern, it is better to reduce the number of sensors such that the subset of selected sensors symmetrically distributes around the source, which in principle would need to know the source location in advance.
969 Anomaly Detection with Training Data in Hyperspectral Imagery J. Liu, Y. Feng, W. Liu, D. Orlando and H. Li In this paper, we investigate the anomaly detection problem for multi-pixel targets in hyperspectral imagery when training data are available.
970 Least-Squares DOA Estimation with an Informed Phase Unwrapping and Full Bandwidth Robustness A. Bohlender, A. Spriet, W. Tirry and N. Madhu In this contribution, we focus on both of the described shortcomings.
971 Joint Blind Calibration and Time-Delay Estimation for Multiband Ranging T. Kazaz, M. Coutino, G. J. M. Janssen and A. van der Veen In this paper, we focus on the problem of blind joint calibration of multiband transceivers and time-delay (TD) estimation of multipath channels.
972 Upgrade Methods for Stratified Sensor Network Self-Calibration M. Larsson, G. Flood, M. Oskarsson and K. ?str?m In the paper new efficient algorithms for solving for the upgrade parameters using minimal data are presented.
973 Audio-Visual Calibration with Polynomial Regression for 2-D Projection Using SVD-PHAT F. Grondin, H. Tang and J. Glass This paper proposes a straightforward 2-D method to spatially calibrate the visual field of a camera with the auditory field of an array microphone by generating and overlaying an acoustic image over an optical image.
974 Weighted Null Vector Initialization and its Application to Phase Retrieval K. Liu, L. Li and L. Wan This paper introduces a simple method, called weighted null vector initialization (WNI), which can be used to compute accurate initialization vectors for solving non-convex phase retrieval.
975 Low-Complexity Accurate Mmwave Positioning for Single-Antenna Users Based on Angle-of-Departure and Adaptive Beamforming A. Fascista, A. Coluccia, H. Wymeersch and G. Seco-Granados Based on this analysis, a low-complexity two-step algorithm with improved localization performance is proposed, which first performs a (coarse) angle of departure estimation and then precodes the down-link signal to introduce beamforming towards the user direction.
976 Accurate Localization of AUV in Motion by Explicit Solution Using Time Delays T. Jia, K. C. Ho, H. Wang and X. Shen A new time delay model that accounts for the motion is proposed for moving AUV localization.
977 Drss-Based Localisation Using Weighted Instrumental Variables and Selective Power Measurement J. Li, K. Dogan?ay, N. H. Nguyen and Y. Wei Law This paper addresses the bias problem when the noise is large by using the method of selective power measurement to maintain correlation between the instrumental variable and data matrix.
978 A Simple and Efficient Iterative Method for Toa Localization Y. Zou and H. Liu This paper develops a simple and efficient method for source localization using signal time-of-arrival (TOA) measurements.
979 Distributed Tracking and Circumnavigation Using Bearing Measurements A. Parayil and J. George This paper is concerned with the problem of bearings based multi-agent circumnavigation of a maneuvering target.
980 Joint Multitarget Tracking and Dynamic Network Localization in the Underwater Domain R. Mendrzik et al. In contrast to commonly-used approaches which split the sensor localization and target tracking into two different sub-problems, we propose a holistic approach for joint localization and tracking.
981 Robust Tdoa Indoor Tracking Using Constrained Measurement Filtering and Grid-Based Filtering R. Huang, J. Tao, L. Yang, Y. Xue and Q. Wu This paper presents an enhanced two-step approach to achieve robust TDOA indoor tracking.
982 Extended Object Tracking Using Hierarchical Truncation Measurement Model with Automotive Radar Y. Xia et al. Motivated by real-world automotive radar measurements that are distributed around object (e.g., vehicles) edges with a certain volume, a novel hierarchical truncated Gaussian measurement model is proposed to resemble the underlying spatial distribution of radar measurements.
983 DOA Tracking Via Signal-Subspace Projector Update J. Zhuang, T. Tan, D. Chen and J. Kang We develop a novel direction-of-arrival (DOA) tracking method in which we directly operate the signal-subspace projector instead of tracking the subspace eigenbasis.
984 Parameter Estimation of In-City Frontal Rainfall Propagation M. Hadar, J. Ostrometzky and H. Messer In this study, we demonstrate that standard signal-level measurements being collected by the network can be used to estimate the movement of an ongoing storm.
985 ML and EM Estimation of Sampling Intervals of Sensor Devices R. Nishimura and Y. Suzuki A precise method of estimating update interval of the register is discussed herein in conjunction with modeling of the problem.
986 Asymptotic Stochastic Analysis of Partially Relaxed DML D. Schenck, X. Mestre and M. Pesavento In this paper, we investigate the outlier production mechanism of the Partially Relaxed Deterministic Maximum Likelihood (PR-DML) Direction-of-Arrival estimator using tools from Random Matrix Theory.
987 Theoretical Performance Bound of Uplink Channel Estimation Accuracy in Massive MIMO A. Osinsky, A. Ivanov and D. Yarotsky In this paper, we present a new performance bound for uplink channel estimation (CE) accuracy in the Massive Multiple Input Multiple Output (MIMO) system.
988 Signal-Aware Broadband DOA Estimation Using Attention Mechanisms W. Mack, U. Bharadwaj, S. Chakrabarty and E. A. P. Habets To obtain a flexible signal-aware DOA estimator, we propose to use binary mask attention with a DNN for multi-source DOA estimation trained with artificial noise.
989 Static Visual Spatial Priors for DoA Estimation P. Swietojanski and O. Miksik In this paper, we propose what to our knowledge is the first multi-modal direction of arrival (DoA) of sound, which uses static visual spatial prior providing an auxiliary information about the environment to suppress some of the false DoA detections.
990 Mirrored Arrays for Direction-of-Arrival Estimation D. Zhu, G. Li and X. Zhang A mirrored array configuration, which consists of an ordinary linear array and a reflector, is proposed.
991 Time Difference of Arrival Estimation from Frequency-Sliding Generalized Cross-Correlations Using Convolutional Neural Networks L. Comanducci, M. Cobos, F. Antonacci and A. Sarti Inspired by deep-learning-based image denoising solutions, we propose in this paper the use of convolutional neural networks (CNNs) to learn the time-delay patterns contained in FS-GCCs extracted in adverse acoustic conditions.
992 Group-Utility Metric for Efficient Sensor Selection and Removal in LCMV Beamformers A. M. Narayanan and A. Bertrand In this paper we derive a generalized expression to efficiently calculate the utility of such groups as a whole, called the group-utility.
993 Accurate Semidefinite Relaxation Method for 3-D Rigid Body Localization Using AOA G. Wang and K. C. Ho This paper addresses the rigid body localization problem using angle-of-arrival measurements. We formulate the problem as a constrained weighted least squares (CWLS) minimization problem with the rotation matrix and position vector as variables, which is a challenging non-convex problem.
994 Cram?r-Rao Bounds for Flaw Localization in Subsampled Multistatic Multichannel Ultrasound Ndt Data E. P?rez, J. Kirchhof, S. Semper, F. Krieg and F. R?mer In this paper, we analyze a compressed sensing scenario in which location parameters of point-like scatterers are estimated from subsampled FMC data.
995 Age of Information with Finite Horizon and Partial Updates D. Ramirez, E. Erkip and H. V. Poor Age of Information with Finite Horizon and Partial Updates
996 Robust Online Mirror Saddle-Point Method for Constrained Resource Allocation E. Tampubolon and H. Boche For this reason, we introduce a new performance measure, whose particular instance is the cumulative positive part of the constraint violations.
997 Real-Time Task Offloading for Large-Scale Mobile Edge Computing Y. Xu, P. Cheng, Z. Chen, M. Ding, Y. Li and B. Vucetic In this paper, we propose a novel index based real-time task offloading policy for an asynchronous large-scale MEC system.
998 Simple Caching Schemes for Non-homogeneous MISO Cache-Aided Communication via Convexity I. Bergel and S. Mohajer We present a novel scheme for cache-aided communication over multiple-input and single output (MISO) cellular networks.
999 Dynamic Resource Optimization and Altitude Selection in Uav-Based Multi-Access Edge Computing F. Costanzo, P. D. Lorenzo and S. Barbarossa The aim of this work is to develop a dynamic optimization strategy to allocate communication and computation resources in a Multi-access Edge Computing (MEC) scenario, where Unmanned Aerial Vehicles (UAVs) act as flying base station platforms endowed with computation capabilities to provide edge cloud services on demand.
1000 Joint Resource Allocation and Routing for Service Function Chaining with In-Subnetwork Processing N. Reyhanian, H. Farmanbar, S. Mohajer and Z. Luo In this paper, we study the joint problem of VNF placement on servers and traffic engineering for a network spanning multiple subnetworks.
1001 Online Channel Estimation for Hybrid Beamforming Architectures J. Fink, R. L. G. Cavalcante and S. Stanczak In this paper, we use variants of the adaptive projected subgradient method to devise online estimation algorithms that exploit temporal correlations of channel samples.
1002 An Optimal Channel Estimation Scheme for Intelligent Reflecting Surfaces Based on a Minimum Variance Unbiased Estimator T. L. Jensen and E. De Carvalho In a wireless system with Intelligent Reflective Surfaces (IRS) containing many passive elements, we consider the problem of channel estimation.
1003 Low-Rank MMWAVE MIMO Channel Estimation in One-Bit Receivers N. J. Myers, K. N. Tran and R. W. Heath In this paper, we develop channel estimation algorithms that exploit the low-rank property of mmWave channels.
1004 Channel Charting: an Euclidean Distance Matrix Completion Perspective P. Agostini, Z. Utkovski and S. Stanczak In this paper, we introduce a correlation matrix distance (CMD) based dissimilarity measure for CC that allows us to group CSI measurements according to their co-linearity.
1005 MMSE-Based Channel Estimation for Hybrid Beamforming Massive MIMO with Correlated Channels J. Mirzaei, F. Sohrabi, R. Adve and S. ShahbazPanahi In this paper, we study the channel estimation problem in microwave correlated massive multiple-input-multiple-output systems with reduced number of radio-frequency chains.
1006 Complex Trainable Ista for Linear and Nonlinear Inverse Problems S. Takabe, T. Wadayama and Y. C. Eldar In this paper, we propose a trainable iterative signal recovery algorithm named complex-field TISTA (C-TISTA) which treats complex-field nonlinear inverse problems.
1007 Conditional Mutual Information Neural Estimator S. Molavipour, G. Bassi and M. Skoglund Motivated by the fact that, in many communication schemes, the achievable transmission rate is determined by a conditional mutual information term, this paper focuses on neural-based estimators for this information-theoretic quantity.
1008 Q-Learning Based Predictive Relay Selection for Optimal Relay Beamforming A. Dimas, K. Diamantaras and A. P. Petropulu This paper considers a scenario of relay beamforming, in which relays collaboratively retransmit the source signal so that they maximize the average signal-to-interference+noise ratio (SINR) at the destination.
1009 Peer To Peer Offloading With Delayed Feedback: An Adversary Bandit Approach M. Yang, H. Zhu, H. Wang, Y. Koucheryavy, K. Samouylov and H. Qian In this paper, we consider a peer computation offloading problem for a fog network with unknown dynamics.
1010 Transferable Policies for Large Scale Wireless Networks with Graph Neural Networks M. Eisen and A. Ribeiro To learn a transferable policy that can generalize to varying and growing networks, we propose the use of so-called random edge graph neural networks (REGNNs).
1011 A Zeroth-Order Learning Algorithm for Ergodic Optimization of Wireless Systems with no Models and no Gradients D. S. Kalogerias, M. Eisen, G. J. Pappas and A. Ribeiro Under a modular stochastic functional optimization framework, we propose a new zeroth-order stochastic primal-dual algorithm for completely data-driven, model-free and gradient-free learning of optimal resource allocation policies for ergodic network optimization.
1012 Joint Sparse Recovery Using Deep Unfolding With Application to Massive Random Access A. P. Sabulal and S. Bhashyam We propose a learning-based joint sparse recovery method for the multiple measurement vector (MMV) problem using deep unfolding.
1013 Learning-Based Content Caching and User Clustering: A Deep Deterministic Policy Gradient Approach K. Chan and F. Chien We propose a new learning structure, termed multiDDPG (MDDPG), that demonstrates better EE performance while providing a comparable CHP to the caching scheme with known content popularity.
1014 Learning-Aided Content Placement in Caching-Enabled fog Computing Systems Using Thompson Sampling J. Zhu, X. Huang and Z. Shao In this paper, we focus on the problem of online content placement with unknown content popularity in caching-enabled fog computing systems, i.e., how to decide and update cached content on resourcelimited edge fog nodes to maximize cache hit rate and minimize switching costs of content update.
1015 Joint Coding and Modulation in the Ultra-Short Blocklength Regime for Bernoulli-Gaussian Impulsive Noise Channels Using Autoencoders K. Vedula, R. Paffenroth and D. R. Brown This paper develops a joint coding and modulation scheme for end-to-end communication system design using an autoencoder architecture in the ultra-short blocklength regime.
1016 Deep Joint Source-Channel Coding for Wireless Image Retrieval M. Jankowski, D. G?nd?z and K. Mikolajczyk Motivated by surveillance applications with wireless cameras or drones, we consider the problem of image retrieval over a wireless channel.
1017 Meta-Learning to Communicate: Fast End-to-End Training for Fading Channels S. Park, O. Simeone and J. Kang In this paper, we propose to obviate the limitations of joint training via meta-learning: Rather than training a common model for all channels, meta-learning finds a common initialization vector that enables fast training on any channel.
1018 Complexity Reduction Methods for Index Modulation Based Dual-Function Radar Communication Systems T. Huang, N. Shlezinger, X. Xu, Y. Liu and Y. C. Eldar In this work we propose schemes for reducing the decoding complexity associated with IM-based DFRC systems.
1019 Equalization of OFDM Waveforms with Insufficient Cyclic Prefix D. Gregoratti and X. Mestre In this paper, a simple equalization strategy for OFDM waveforms is proposed that specifically targets the case where the cyclic prefix is insufficient to span the whole channel duration.
1020 Faster-Than-Nyquist Signaling Via Spatiotemporal Symbol-Level Precoding for Multi-User MISO Redundant Transmissions W. A. Martins, D. Spano, S. Chatzinotas and B. Ottersten We propose a framework for redundant block-based symbol-level precoders enabling the trade-off between constructive and destructive multi-user and interblock interference (IBI) effects at the singleantenna user terminals.
1021 Optimized Single Carrier Transceiver for Future Sub-TeraHertz Applications S. Bica?s, J. Dor?, G. Goug?on and Y. Corre In this paper, we address the design of a single carrier transceiver resilient to phase noise.
1022 Power Spectrum Optimization for Capacity of the Extended Spectrum Hybrid Fiber Coax Network R. Strobel and T. Hewavithana This work presents a new method of spectrum optimization for the coax channel up to 3GHz, taking transmitter distortion into account in the optimization.
1023 A Low-Latency Successive Cancellation Hybrid Decoder for Convolutional Polar Codes Y. Wang et al. With the observation that SCL and SCF decoding are similar at giving more chances to inspect possible codewords simultaneously or sequentially, a novel hybrid decoder is proposed in this paper, which essentially combines the ideas of SCF and SCL decoders.
1024 Near Capacity RCQD Constellations for PAPR Reduction of OFDM Systems T. Arbi, I. Nasr and B. Geller We investigate an optimized blind SeLected Mapping (SLM) algorithm to reduce the Peak-to-Average Power Ratio (PAPR) for Orthogonal Frequency Division Multiplexing (OFDM) systems with Signal Space Diversity (SSD).
1025 Fully Pipelined Iteration Unrolled Decoders the Road to TB/S Turbo Decoding S. Weithoffer, R. Klaimi, C. A. Nour, N. Wehn and C. Douillard In this paper, we present recent findings on the implementation of ultra-high throughput Turbo decoders.
1026 Zero-Crossing Precoding with Maximum Distance to the Decision Threshold for Channels with 1-Bit Quantization and Oversampling D. M. V. Melo, L. T. N. Landau and R. C. de Lamare In this study, we propose a novel waveform for bandlimited channels with 1-bit quantization and oversampling at the receivers.
1027 Achieving Fully-Digital Performance by Hybrid Analog/Digital Beamforming in Wide-Band Massive-Mimo Systems A. Morsali and B. Champagne In this paper, we study the realization of any given fullydigital precoder (FDP) by hybrid analog/digital precoding (HADP) in wide-band mmWave systems.
1028 Energy-Efficient Bit Allocation for Resolution-Adaptive ADC in Multiuser Large-Scale MIMO Systems: Global Optimality K. Nguyen, Q. Vu, L. Tran and M. Juntti We consider uplink multiuser wireless communications systems, where the base station (BS) receiver is equipped with a large-scale antenna array and resolution adaptive analog-to-digital converters (ADCs).
1029 Generalized Spatial Modulation for Wireless Terabits Systems Under Sub-THZ Channel With RF Impairments M. Saad, F. Bader, A. C. Al Ghouwayel, H. Hijazi, N. Bouhel and J. Palicot The obtained results reveal that QPSK-GSM system is the best combination compared to GSM systems with any other Mary modulation scheme (e.g. PSK, DPSK, QAM, PAM).
1030 Efficient Techniques For in-band System Information Broadcast in Multi-Cell Massive Mimo J. Jayachandran, K. Biswas, S. K. Mohammed and E. G. Larsson In this paper we consider joint beamforming of data to scheduled terminals (STs) and broadcast of system information (SI) to idle terminals (ITs) on the same time-frequency resource in multi-cell multi-user massive MIMO systems.
1031 Optimal Design of Energy-Efficient Cell-Free Massive Mimo: Joint Power Allocation and Load Balancing T. V. Chien, E. Bj?rnson and E. G. Larsson To find a flexible and energy-efficient implementation, we minimize the total power consumption at the APs in the downlink, considering both the hardware and transmit powers, where APs can be turned off.
1032 Large-Scale Fading Precoding for Maximizing the Product of SINRs ?. T. Demir and E. Bj?rnson This paper considers the large-scale fading precoding design for mitigating the pilot contamination in the downlink of multi-cell massive MIMO (multiple-input multiple-output) systems.
1033 Proximal Distance Algorithm for Nonconvex QCQP with Beamforming Applications Q. Li, Y. Liu, M. Shao and W. Ma In light of this, this work aims at developing an efficient approach to the QCQP without the above mentioned drawbacks.
1034 Cloud-Driven Multi-Way Multiple-Antenna Relay Systems: Best-User-Link Selection and Joint Mmse Detection F. L. Duarte and R. C. de Lamare In this work, we present a cloud-driven uplink framework for multiway multiple-antenna relay systems which facilitates joint linear Minimum Mean Square Error (MMSE) symbol detection in the cloud and where users are selected to simultaneously transmit to each other aided by relays.
1035 A Complexity Efficient DMT-Optimal Tree Pruning Based Sphere Decoding M. Neinavaie, M. Derakhtian and S. A. Vorobyov We present a diversity multiplexing tradeoff (DMT) optimal tree pruning sphere decoding algorithm which visits merely a single branch of the search tree of the sphere decoding (SD) algorithm, while maintaining the DMT optimality at high signal to noise ratio (SNR) regime.
1036 A Model-Free Approach to Distributed Transmit Beamforming J. George, C. T. Yilmaz, A. Parayil and A. Chakrabortty This paper presents a model-free solution to distributed transmit beamforming using mobile agents.
1037 Intelligent Reflecting Surface for Massive Device Connectivity: Joint Activity Detection and Channel Estimation S. Xia and Y. Shi In this paper, we formulate the IRS-related activity detection and channel estimation problem as sparse matrix factorization, matrix completion and multiple measurement vector problem and, we propose a three-stage framework based on the approximate message passing.
1038 Channel Covariance Estimation in Multiuser Massive Mimo Systems with an Approach Based on Infinite Dimensional Hilbert Spaces R. L. G. Cavalcante and S. Stanczak We propose a novel algorithm to estimate the channel covariance matrix of a desired user in multiuser massive MIMO systems.
1039 Eliminating Out-Of-Cell Interference in Cellular Massive Mimo with a Single Additional Transceiver U. Erez and A. Leshem The present work considers a line-of-sight channel model. It is shown that given a sufficiently large antenna array, it suffices that the number of transmit/receive chains exceeds the number of desired users by one in order to reduce the interference to any desired level by judiciously selecting the antenna elements.
1040 Favorable Propagation and Linear Multiuser Detection for Distributed Antenna Systems R. Gholami, L. Cottatellucci and D. Slock In this paper, we consider a wireless network with transmit and receive antennas distributed according to homogeneous point processes.
1041 Distributed Non-Orthogonal Pilot Design for Multi-Cell Massive Mimo Systems Y. Wu, S. Med and Y. Gu In this work, a distributed non-orthogonal pilot design approach is proposed to tackle the pilot contamination problem in multi-cell massive multiple input multiple output (MIMO) systems.
1042 Distributed Detection of Sparse Signals with 1-Bit Data in Two-Level Two-Degree Tree-Structured Sensor Networks C. Li, G. Li and P. K. Varshney In this paper, we present a new detector for the detection of sparse stochastic signals using 1-bit data in two-level two- degree tree-structured sensor networks (2L-2D TSNs).
1043 Objective Bayesian Detection Under Spatially Correlated Gaussian Observations for Multi-Antenna Cognitive Radio Network M. H. Al-Ali and K. C. Ho This paper develops an objective Bayesian detector for asserting the presence of primary user (PU) signal buried in additive noise/interference using a sequence of complex vector samples from a multi-antenna spectrum sensing system.
1044 A Gated Hypernet Decoder for Polar Codes E. Nachmani and L. Wolf In this work, we demonstrate how hypernet-works can be applied to decode polar codes by employing a new formalization of the polar belief propagation decoding scheme.
1045 Weighted Gradient Coding with Leverage Score Sampling N. Charalambides, M. Pilanci and A. O. Hero We present a novel weighted leverage score approach, that achieves improved performance for distributed gradient coding by utilizing an importance sampling.
1046 Low-Complexity 5g Slam with CKF-PHD Filter H. Kim, K. Granstr?m, S. Kim and H. Wymeersch This paper proposes a new implementation method for the 5G SLAM using message passing (MP) and the cubature Kalman filter (CKF).
1047 The Effect of Power Allocation on Visible Light Communication Using Commercial Phosphor-Converted Led Lamp for Indirect Illumination A. A. Dowhuszko, M. Ilter, P. Pinho and J. H?m?l?inen In this paper, we estimate the data rate that is feasible with VLC when indirect illumination is used.
1048 Robust Transmission Over Channels with Channel Uncertainty: an Algorithmic Perspective H. Boche, R. F. Schaefer and H. Vincent Poor This paper approaches this question from a fundamental, algorithmic point of view to study whether or not such optimal schemes can be found algorithmically in principle (without putting any constraints on the computational complexity of such algorithms).
1049 Deep Joint Source-Channel Coding of Images with Feedback D. B. Kurka and D. G?nd?z We consider wireless transmission of images in the presence of channel output feedback, by introducing an autoencoder-based deep joint source-channel coding (JSCC) scheme.
1050 A Learning Approach to Cooperative Communication System Design Y. Lu, P. Cheng, Z. Chen, W. H. Mow and Y. Li We present in this paper a Neural Network (NN)-based autoencoder (AE) approach to optimize its design.
1051 A Stacked-Autoencoder Based End-to-End Learning Framework for Decode-and-Forward Relay Networks A. Gupta and M. Sellathurai In this work, we study an end-to-end deep learning (DL)based constellation design for decode-and-forward (DF) relay network.
1052 A New Sampling Scheme for Distributed Blind Spectrum Sensing Using Energy Detectors T. Wang, F. Chien and C. Hsieh In this paper, we study the problem of blind spectrum sensing by exploring signal sampling at each cognitive radio (CR) in a distributed cognitive radio network.
1053 On Throughput of Millimeterwave Mimo Systems with Lowresolution ADCS A. Khalili, S. Shahsavari, F. Shirani, E. Erkip and Y. C. Eldar In this paper, a receiver with low resolution ADCs based on adaptive thresholds is considered in downlink mmWave communications in which the channel state information is not known a-priori and acquired through channel estimation.
1054 Reliable and Secure Transmission for Future Networks Y. Hua This paper introduces a novel physical layer encryption method called randomized reciprocal channel modulation (RRCM) for reliable and secure transmission of information against eavesdropper (Eve) with any number of antennas and any noise level.
1055 On Polar Coding For Finite Blocklength Secret Key Generation Over Wireless Channels H. Hentil?, Y. Y. Shkel, V. Koivunen and H. V. Poor We consider the problem of secret key generation from correlated Gaussian random variables in the finite blocklength regime.
1056 An Enhanced Decoding Algorithm for Coded Compressed Sensing V. K. Amalladinne, J. Chamberland and K. R. Narayanan This article introduces an enhanced decoding algorithm for coded compressed sensing where fragment recovery and the stitching process are executed in tandem, passing information between them.
1057 Optimal Window Design for W-OFDM K. Hussain and R. L?pez-Valcarce We present an optimal window design for windowed-OFDM minimizing OBR within a given frequency region.
1058 Computability of the Peak Value of Bandlimited Signals H. Boche and U. J. M?nich In this paper we study the peak value problem, i.e., the task of computing the peak value of a bandlimited signal from its samples.
1059 Joint Scheduling and Beamforming for Delay Sensitive Traffic with Priorities and Deadlines I. Hadar and A. Leshem In this paper we propose a simple scheduling algorithm which takes priorities and deadlines into account and allocates users dynamically to resource blocks and spatial beams according to the Tomlinson-Harashima precoder.
1060 Robust Hybrid Precoding For Interference Exploitation in Massive Mimo Systems G. Hegde, C. Masouros and M. Pesavento In this paper, we consider a multiuser massive MIMO system with hybrid analog-digital precoding architecture.
1061 State-Space Gaussian Process for Drift Estimation in Stochastic Differential Equations Z. Zhao, F. Tronarp, R. Hostettler and S. S?rkk? We propose to formulate this as a non-parametric Gaussian process regression problem and use an Ito-Taylor expansion for approximating the SDE.
1062 Computing Hilbert Transform and Spectral Factorization for Signal Spaces of Smooth Functions H. Boche and V. Pohl This paper characterizes for both operations precisely those signal spaces of differentiable functions for which such an effective control of the approximation error is possible.
1063 M-Estimators of Scatter with Eigenvalue Shrinkage E. Ollila, D. P. Palomar and F. Pascal In this paper, a more general approach is considered in which the SCM is replaced by an M-estimator of scatter matrix and a fully automatic data adaptive method to compute the optimal shrinkage parameter with minimum mean squared error is proposed.
1064 A Multitaper Reassigned Spectrogram for Increased Time-Frequency Localization Precision M. Sandsten, I. Reinhold and R. Anderson In this paper we propose a multitaper reassignment (mtRS) method for estimation of time- and frequency locations of Gaussian envelope transients.
1065 Stochastic Ml Estimation for Hyperspectral Unmixing Under Endmember Variability and Nonlinear Models Y. Li, R. Wu and W. Ma In this study we introduce a probabilistic approach for HU under two different models, namely, the normal composition model for modeling spatial endmember variability and the generalized bilinear model for modeling multi-path nonlinear effects.
1066 Robust Phase Retrieval with Outliers X. Jiang, H. C. So and X. Liu An outlier-resistance phase retrieval algorithm based on alternating direction method of multipliers (ADMM) is devised in this paper.
1067 Node-Asynchronous Spectral Clustering On Directed Graphs O. Teke and P. P. Vaidyanathan Assuming that the graph operator has eigenvalue 1 and the input signal satisfies a certain condition (which ensures the existence of fixed points), this study presents the necessary and sufficient condition for the mean-squared convergence of the graph signal.
1068 Estimating Centrality Blindly From Low-Pass Filtered Graph Signals Y. He and H. Wai To remedy, we propose a robust blind centrality estimation method which substantially improves the centrality estimation performance.
1069 Blind Inference of Centrality Rankings from Graph Signals T. M. Roddenberry and S. Segarra We study the blind centrality ranking problem, where our goal is to infer the eigenvector centrality ranking of nodes solely from nodal observations, i.e., without information about the topology of the network.
1070 A Low-Dimensionality Method for Data-Driven Graph Learning L. Stankovic, M. Dakovic, D. Mandic, M. Brajovic, B. Scalzo-Dees and A. G. Constantinides In this paper, we propose a numerically efficient method for estimating of the normalized Laplacian through its eigenvalues estimation and by promoting its sparsity.
1071 Metric Representations of Networks: A Uniqueness Result S. Segarra, T. M. Roddenberry, F. M?moli and A. Ribeiro In this paper, we consider the problem of projecting networks onto metric spaces.
1072 On The Stability of Polynomial Spectral Graph Filters H. Kenlay, D. Thanou and X. Dong In this work, we first prove that polynomial graph filters are stable with respect to the change in the normalised graph Laplacian matrix. We then show empirically that properties of a structural perturbation, specifically the relative locality of the edges removed in a binary graph, effect the change in the normalised graph Laplacian.
1073 On Cram?r-Rao Lower Bounds with Random Equality Constraints C. Pr?vost, E. Chaumette, K. Usevich, D. Brie and P. Comon In this communication, we introduce a new constrained Cram?r-Rao- like bound for observations where the probability density function (p.d.f.) parameterized by unknown deterministic parameters results from the marginalization of a joint p.d.f. depending on random variables as well.
1074 On Harmonic Approximations of Inharmonic Signals F. Elvander, J. Ding and A. Jakobsson In this work, we present the misspecified Gaussian Cram?r-Rao lower bound for the parameters of a harmonic signal, or pitch, when signal measurements are collected from an almost, but not quite, harmonic model.
1075 A General Test for the Linear Structure of Covariance Matrices of Gaussian Populations Y. Xiao, D. Ram?rez and P. J. Schreier We study the problem under the Gaussian assumption and derive the generalized likelihood ratio test (GLRT).
1076 Sequential Joint Detection and Estimation with an Application to Joint Symbol Decoding and Noise Power Estimation D. Reinhard, M. Fau? and A. M. Zoubir The underlying constrained problem is first converted to an unconstrained problem and then reduced to an optimal stopping problem, whose solution is characterized by a non-linear Bellman equation.
1077 A Linear Time Partitioning Algorithm for Frequency Weighted Impurity Functions T. Nguyen and T. Nguyen In this paper, we propose a heuristic, efficient (linear time) algorithm for finding the minimum impurity for a broader class of impurity functions which includes popular impurities such as Gini index and entropy.
1078 Finite Sample Deviation and Variance Bounds for First Order Autoregressive Processes R. A. Gonz?lez and C. R. Rojas In this paper, we study finite-sample properties of the least squares estimator in first order autoregressive processes.
1079 Balancing Rates and Variance via Adaptive Batch-Sizes in First-Order Stochastic Optimization Z. Gao, A. Koppel and A. Ribeiro In this work, we seek to balance the fact that attenuating step-sizes is required for exact asymptotic convergence with the fact that larger constant step-sizes learn faster in finite time up to an error.
1080 A Greedy Sparse Approximation Algorithm Based On L1-Norm Selection Rules R. B. Mhenni, S. Bourguignon and J. Idier We propose a new greedy sparse approximation algorithm, called SLS for Single L1 Selection, that addresses a least squares optimization problem under a cardinality constraint.
1081 Exact Sparse Nonnegative Least Squares N. Nadisic, A. Vandaele, N. Gillis and J. E. Cohen We propose a novel approach to solve exactly the sparse nonnegative least squares problem, under hard l0 sparsity constraints.
1082 Epigraphical Reformulation for Non-Proximable Mixed Norms S. Kyochi, S. Ono and I. Selesnick This paper proposes an epigraphical reformulation (ER) technique for non-proximable mixed norm regularization.
1083 Forward-Backward Splitting for Optimal Transport Based Problems G. Ortiz-Jim?nez, M. El Gheche, E. Simou, H. P. Maretic and P. Frossard To overcome this limitation, we devise a general forward-backward splitting algorithm based on Bregman distances for solving a wide range of optimization problems involving a differentiable function with Lipschitz-continuous gradient and a doubly stochastic constraint.
1084 SSGD: Sparsity-Promoting Stochastic Gradient Descent Algorithm for Unbiased Dnn Pruning C. Lee, I. Fedorov, B. D. Rao and H. Garudadri In this paper, we explore the application of iterative reweighting methods popular in SSR to learning efficient DNNs.
1085 Low-Rank Tensor Ring Model for Completing Missing Visual Data M. S. Asif and A. Prater-Bennette In this paper, we present an algorithm for recovering low-rank tensors from massively under-sampled or missing data.
1086 Sequential Semi-Orthogonal Multi-Level NMF with Negative Residual Reduction for Network Embedding R. Hashimoto and H. Kasai To alleviate this shortcoming, this paper presents a proposal of a sequential semi-orthogonal NMF with negative residual reduction for the network embedding (SSO-NRR-NMF).
1087 A Fast Proximal Point Algorithm for Generalized Graph Laplacian Learning Z. Deng and A. M. So In this paper, we focus on the problem of learning the generalized graph Lapla-cian (GGL) and propose an efficient algorithm to solve it.
1088 Reconstruction of Fri Signals Using Deep Neural Network Approaches V. C. H. Leung, J. Huang and P. L. Dragotti In this paper, we aim to alleviate the subspace swap problem and investigate alternative approaches including directly estimating FRI parameters using deep neural networks and utilising deep neural networks as denoisers to reduce the noise in the samples.
1089 Adaptive Prediction of Financial Time-Series for Decision-Making Using A Tensorial Aggregation Approach B. S. C. Campello, L. T. Duarte and J. M. T. Romano More precisely, we introduce a novel approach in which the data taken into account in the decision process is modeled as a tensor.
1090 Data Selection Kernel Conjugate Gradient Algorithm P. S. R. Diniz, J. O. Ferreira, M. O. K. Mendon?a and T. N. Ferreira In this paper, we propose the data selection kernel conjugate gradient (DS-KCG) algorithm, which is capable of classifying whether the currently available data brings sufficient innovation to update the filter coefficients.
1091 Normalized Least-Mean-Square Algorithms with Minimax Concave Penalty H. Kaneko and M. Yukawa We propose a novel problem formulation for sparsity-aware adaptive filtering based on the nonconvex minimax concave (MC) penalty, aiming to obtain a sparse solution with small estimation bias.
1092 Steepening Squared Error Function Facilitates Online Adaptation of Gaussian Scales M. Takizawa and M. Yukawa In this paper, we propose steepening the cost function by adding a squared distance function from the instantaneously-optimal scale.
1093 Indoor Altitude Estimation of Unmanned Aerial Vehicles Using a Bank of Kalman Filters L. Yang, H. Wang, Y. El-Laham, J. I. L. Fonte, D. T. P?rez and M. F. Bugallo In this work, we propose a novel strategy for tackling the altitude estimation problem that utilizes multiple model adaptive estimation (MMAE), where the candidate models correspond to four scenarios: no obstacles above and below the UAV; obstacles above the UAV; obstacles below the UAV; and obstacles above and below the UAV.
1094 Underwater Tracking Based on the Sum-Product Algorithm Enhanced by a Neural Network Detections Classifier G. Soldi et al. This paper describes a multiobject tracking framework based on the sum-product algorithm that exploits the information provided by a convolutional neural network that classifies the LFAS detections.
1095 Feature Affine Projection Algorithms H. Yazdanpanah In this paper, we propose the Feature Affine Projection (F-AP) algorithm to reveal hidden sparsity in unknown systems.
1096 Approximate Bayesian Computation with the Sliced-Wasserstein Distance K. Nadjahi, V. D. Bortoli, A. Durmus, R. Badeau and U. Simsekli We propose a new ABC technique, called Sliced-Wasserstein ABC and based on the Sliced-Wasserstein distance, which has better computational and statistical properties.
1097 Enhanced Mixture Population Monte Carlo Via Stochastic Optimization and Markov Chain Monte Carlo Sampling Y. El-Laham, P. M. Djuric and M. F. Bugallo In this work, we introduce a novel M-PMC algorithm that optimizes the parameters of a mixture proposal distribution, where parameter updates are resolved via stochastic optimization instead of EM.
1098 Better Safe Than Sorry: Risk-Aware Nonlinear Bayesian Estimation D. S. Kalogerias, L. F. O. Chamon, G. J. Pappas and A. Ribeiro To address this issue, we introduce a new risk-aware MMSE formulation which trades between mean performance and risk by explicitly constraining the expected predictive variance of the involved squared error.
1099 Particle Filtering on the Complex Stiefel Manifold with Application to Subspace Tracking C. J. Bordin and M. G. S. Bruno In this paper, we extend previous particle filtering methods whose states were constrained to the (real) Stiefel manifold to the complex case.
1100 Bayesian Multiple Change-Point Detection with Limited Communication T. Halme, E. Nitzan, H. V. Poor and V. Koivunen In this paper, we consider Bayesian multiple changepoint detection using a sensor network in which a fusion center can receive a data stream from each sensor.
1101 What did your adversary believe? Optimal Filtering and Smoothing in Counter-Adversarial Autonomous Systems R. Mattila, I. Louren?o, V. Krishnamurthy, C. R. Rojas and B. Wahlberg We consider fixed-interval smoothing problems for counter-adversarial autonomous systems.
1102 Robust Parameter Estimation of Contaminated Damped Exponentials Y. Xie, D. Liu, H. Mansour and P. T. Boufounos In this paper we aim to estimate parameters of damped exponentials from contaminated signal, i.e., a mixture of damped exponentials, random Gaussian noise, and spike interference.
1103 Computation of “Best” Interpolants in the Lp Sense P. Bohra and M. Unser For this continuous-domain problem, we propose an exact discretization scheme that restricts the search space to quadratic splines with knots on an uniform grid.
1104 Fast Block-Sparse Estimation for Vector Networks Z. Yue, P. Sundaram and V. Solo Here we address key computational bottlenecks and develop a new algorithm which is much faster and has massively reduced requirements on matrix conditioning.
1105 Relative Cost Based Model Selection for Sparse High-Dimensional Linear Regression Models P. B. Gohain and M. Jansson In this paper, we propose a novel model selection method named multi-beta-test (MBT) for the sparse high-dimensional linear regression model.
1106 Cumulant Slice Reconstruction from Compressive Measurements and Its Application to Line Spectrum Estimation Y. Wang and Z. Tian To overcome these challenges, this paper develops a novel compressive cumulant slice sensing (CCSS) method that aims to efficiently reconstructing the 1D diagonal slice of higher-order cumulants under the compressive sensing framework.
1107 Adaptation and Learning in Multi-Task Decision Systems S. Marano and A. H. Sayed Elaborating on previous works on single-task networks engaged in decision problems, here we consider the multi-task version in the challenging scenario where the state of nature may change arbitrarily.
1108 Graph Metric Learning via Gershgorin Disc Alignment C. Yang, G. Cheung and W. Hu We propose a general projection-free metric learning framework, where the minimization objective ${\min _{{\mathbf{M}} \in \mathcal{S}}}Q({\mathbf{M}})$ is a convex differentiable function of the metric matrix M, and M resides in the set S of generalized graph Laplacian matrices for connected graphs with positive edge weights and node degrees.
1109 Learning Graph Influence from Social Interactions V. Matta, V. Bordignon, A. Santos and A. H. Sayed Specifically, from observations of the stream of beliefs at certain agents, we would like to examine whether it is possible to learn the strength of the connections (influences) from sending components in the network to these receiving agents.
1110 Social Learning with Partial Information Sharing V. Bordignon, V. Matta and A. H. Sayed This work studies the learning abilities of agents sharing partial beliefs over social networks.
1111 Non-parametric Community Change-points Detection in Streaming Graph Signals A. Ferrari and C. Richard The aim of this paper is to address this challenge when anomalies activate unknown groups of nodes in a network.
1112 Spatial Gating Strategies for Graph Recurrent Neural Networks L. Ruiz, F. Gama and A. Ribeiro To address this, we propose two spatial gating strategies for GRNNs leveraging the node and edge structure of the graph.
1113 Learning connectivity and higher-order interactions in radial distribution grids Q. Yang, M. Coutino, G. Wang, G. B. Giannakis and G. Leus Establishing a connection between the celebrated exact distribution flow equations and the so-called self-driven graph Volterra model, this paper puts forth a nonlinear topology identification algorithm, that is able to reveal both the edge connections as well as their higher-order interactions.
1114 Semi-Supervised Learning of Processes Over Multi-Relational Graphs Q. Lu, V. N. Ioannidis and G. B. Giannakis Towards this end, a structured dynamical model is introduced to capture the spatio-temporal nature of dynamic graph processes, and incorporate contributions from multiple relations of the graph in a probabilistic fashion.
1115 Recursive Prediction of Graph Signals With Incoming Nodes A. Venkitaraman, S. Chatterjee and B. Wahlberg Keeping this premise in mind, we propose a method to recursively obtain the optimal prediction or regression coefficients for the recently proposed Linear Regression over Graphs (LRG), as the graph expands with incoming nodes.
1116 Learning Signed Graphs from Data G. Matz and T. Dittrich In this paper, we propose a conceptually simple and flexible approach to signed graph learning via signed smoothness metrics.
1117 Forecasting Multi-Dimensional Processes Over Graphs A. Natali, E. Isufi and G. Leus To tackle this issue, we devise a new framework and propose new methodologies based on the graph vector autoregressive model.
1118 Distributed Quantization for Sparse Time Sequences A. Cohen, N. Shlezinger, S. Salamatian, Y. C. Eldar and M. M?dard In this work we propose a distributed quantization scheme for representing a set of sparse time sequences acquired using conventional scalar ADCs.
1119 A Time-Based Sampling Framework for Finite-Rate-of-Innovation Signals S. Rudresh, A. J. Kamath and C. Sekhar Seelamantula In this paper, we consider the problem of sampling and perfect reconstruction of periodic finite-rate-of-innovation (FRI) signals using crossing-time-encoding machine (C-TEM) and integrate-and-fire TEM (IF-TEM).
1120 Effective Approximation of Bandlimited Signals and Their Samples H. Boche and U. J. M?nich In this paper we analyze if and how this transition affects the computability of the signal.
1121 Receiver Design and AGC optimization with Self Interference Induced Saturation C. K. Sheemar and D. Slock This leads to missing samples which we propose to reconstruct under the assumptions that the receive signal of interest is a low pass bandlimited signal with known spectrum (mask), oversampling and (perfect) digital SIC after the ADC.
1122 D-SLAM: Diffusion Source Localization and Trajectory Mapping R. Alexandru, T. Blu and P. L. Dragotti Within this framework, we propose a method for localizing sources of unknown amplitudes, and known activation times.
1123 Triggerless Random Interleaved Sampling M. W. Rupniewski We show that periodic signals are determined by the probabilistic distribution of their sample sequences, provided that the inter-sequence offsets are uniformly distributed (in a probabilistic sense).
1124 The Fractional Quaternion Fourier Number Transform L. C. da Silva, J. R. de Oliveira Neto and J. B. Lima In this paper, we define a fractional version of the quaternion Fourier number transform (QFNT).
1125 Short and Squeezed: Accelerating the Computation of Antisparse Representations with Safe Squeezing C. Elvira and C. Herzet In this paper, we propose a new methodology, coined “safe squeezing”, accelerating the computation of antisparse representations.
1126 Decentralized expected consistent signal recovery for quantization Measurements C. Wang, C. Wen, S. Tsai and S. Jin In this study, we develop a novel decentralized architecture by leveraging the core framework of GEC-SR called “deGEC-SR.”
1127 Lie Group State Estimation via Optimal Transport Z. Wang and V. Solo Here we overcome these problems by managing the geometry with the Cayley transform and particle depletion with optimal transport.
1128 Smoothing Graph Signals via Random Spanning Forests Y. Y. Pilavci, P. Amblard, S. Barthelm? and N. Tremblay Another facet of the elegant link between random processes on graphs and Laplacian-based numerical linear algebra is uncovered: based on random spanning forests, novel Monte-Carlo estimators for graph signal smoothing are proposed.
1129 Diagonalizable Shift and Filters for Directed Graphs Based on the Jordan-Chevalley Decomposition P. Misiakos, C. Wendler and M. P?schel In this paper, we propose to replace a given adjacency shift A by a diagonalizable shift AD obtained via the Jordan-Chevalley decomposition.
1130 Gaussian Processes Over Graphs A. Venkitaraman, S. Chatterjee and P. Handel In this work, we consider the development of a stochastic or Bayesian variant of KRG.
1131 Blind Source Separation of Graph Signals J. Miettinen, S. A. Vorobyov and E. Ollila We propose BSS of graph signals which uses the prior information presented by the signal graph together with nonGaussianity.
1132 Graphical Evolutionary Game Theoretic Analysis of Super Users in Information Diffusion Y. Li, Y. Li, H. Hu, H. V. Zhao and Y. Chen To address the existence of “super users” in social networks who have higher social status and potentially larger influence, we propose a graphical evolutionary game theoretic framework to investigate the impact of such super users and their strategy update rules on information propagation.
1133 Gradient-Based Algorithm with Spatial Regularization for Optimal Sensor Placement F. Ghayem, B. Rivet, C. Jutten and R. C. Farias In this paper, we are interested in optimal sensor placement for signal extraction.
1134 The Graphon Fourier Transform L. Ruiz, L. F. O. Chamon and A. Ribeiro We define graphon signals and introduce the Graphon Fourier Transform (WFT), to which the Graph Fourier Transform (GFT) is shown to converge.
1135 Learning Product Graphs from Multidomain Signals S. K. Kadambari and S. Prabhakar Chepuri In this paper, we focus on learning the underlying product graph structure from multidomain training data.
1136 Graph Vertex Sampling with Arbitrary Graph Signal Hilbert Spaces B. Girault, A. Ortega and S. S. Narayayan In this context, we propose to extend sampling set selection based on spectral proxies to arbitrary Hilbert spaces of graph signals.
1137 Estimation of Information in Parallel Gaussian Channels via Model Order Selection C. A. L?pez, F. de Cabrera and J. Riba We study the problem of estimating the overall mutual information in M independent parallel discrete-time memory-less Gaussian channels from N independent data sample pairs per channel (inputs and outputs).
1138 Generalized Graph Spectral Sampling with Stochastic Priors J. Hara, Y. Tanaka and Y. C. Eldar In this paper, we assume the graph signals are modeled by graph wide sense stationarity (GWSS), which is an extension of WSS for standard time domain signals.
1139 Anomalydae: Dual Autoencoder for Anomaly Detection on Attributed Networks H. Fan, F. Zhang and Z. Li In this paper, we propose a deep joint representation learning framework for anomaly detection through a dual autoencoder (AnomalyDAE), which captures the complex interactions between network structure and node attribute for high-quality embeddings.
1140 On The Degrees Of Freedom in Total Variation Minimization F. Xue and T. Blu Considering the total-variation (TV) regularization, we present a theoretical study of the DOF in Stein?s unbiased risk estimate (SURE), under a very mild assumption.
1141 Atomic Norm Denoising In Blind Two-Dimensional Super-Resolution M. A. Suliman and W. Dai In this work, we develop a new framework for denoising in blind two-dimensional (2D) super-resolution that recovers a set of 2D continuous parameters as well as unknown waveforms from noisy samples.
1142 Dynamic Channel Pruning For Correlation Filter Based Object Tracking G. Y. Gopal and M. A. Amer To mitigate this problem, we propose a method for dynamic channel pruning through online (i.e., at every frame) learning of channel weights.
1143 Positive Semidefinite Matrix Factorization: A Link to Phase Retrieval And A Block Gradient Algorithm D. Lahat and C. F?votte In this paper, we show, for the first time, a link between PS-DMF and the problem of matrix recovery from phaseless measurements, which includes phase retrieval.
1144 Realizability of Planar Point Embeddings from Angle Measurements F. D?mbgen, M. El Helou and A. Scholefield This paper is concerned with the theory of localization from inner-angle measurements.
1145 Sparse Recovery with Non-Linear Fourier Features A. ?z?elikkale Motivated by this success, this article focuses on a sparse non-linear Fourier feature (NFF) model.
1146 Efficient Super-Resolution Two-Dimensional Harmonic Retrieval Via Enhanced Low-Rank Structured Covariance Reconstruction Y. Wang, Y. Zhang, Z. Tian, G. Leus and G. Zhang This paper develops an enhanced low-rank structured covariance reconstruction (LRSCR) method based on the decoupled atomic norm minimization (D-ANM), for super-resolution two-dimensional (2D) harmonic retrieval with multiple measurement vectors.
1147 Effect of Undersampling on Non-Negative Blind Deconvolution with Autoregressive Filters P. Sarangi, M. C. H?c?menoglu and P. Pal Our objective is to understand if it is possible to recover both the signal and the kernel from downsampled measurements of their convolution.
1148 Manifold Gradient Descent Solves Multi-Channel Sparse Blind Deconvolution Provably and Efficiently L. Shi and Y. Chi We propose a novel approach based on nonconvex optimization over the sphere manifold by minimizing a smooth surrogate of the sparsity-promoting loss function.
1149 Sparse Branch and Bound for Exact Optimization of L0-Norm Penalized Least Squares R. B. Mhenni, S. Bourguignon, M. Mongeau, J. Ninin and H. Carfantan We propose a global optimization approach to solve l0-norm penalized least-squares problems, using a dedicated branch-and-bound methodology.
1150 A Proximal Dual Consensus Method for Linearly Coupled Multi-Agent Non-Convex Optimization J. Zhang, S. Ge, T. Chang and Z. Luo In this paper, we propose such a method, called the proximal dual consensus (PDC) method, that combines a proximal technique and the dual consensus method.
1151 A Penalty Alternating Direction Method of Multipliers for Decentralized Composite Optimization J. Zhang, A. M. So and Q. Ling This paper proposes a penalty alternating direction method of multipliers (ADMM) to minimize the summation of convex composite functions over a decentralized network.
1152 Wirtinger Flow Algorithms for Phase Retrieval from Binary Measurements V. Kishore and C. S. Seelamantula We consider the problem of Binary Phase Retrieval, wherein we attempt to recover signals from their quadratic measurements, which are further encoded as +1 or -1 depending on whether they exceed a threshold or not.
1153 Decentralized Min-Max Optimization: Formulations, Algorithms and Applications in Network Poisoning Attack I. Tsaknakis, M. Hong and S. Liu This paper discusses formulations and algorithms which allow a number of agents to collectively solve problems involving both (non-convex) minimization and (concave) maximization operations.
1154 An Efficient Augmented Lagrangian-Based Method for Linear Equality-Constrained Lasso Z. Deng, M. Yue and A. M. So In this paper, we demonstrate how the recently developed semis-mooth Newton-based augmented Lagrangian framework can be extended to solve a linear equality-constrained Lasso model.
1155 Control of Linear Dynamical Systems Using Sparse Inputs C. Sriram, G. Joseph and C. R. Murthy In this work, we consider control of linear dynamical systems using sparse inputs.
1156 Decentralized Stochastic Non-Convex Optimization over Weakly Connected Time-Varying Digraphs S. Lu and C. W. Wu In this paper, we consider decentralized stochastic non-convex optimization over a class of weakly connected digraphs.
1157 Paco and Paco-Dct: Patch Consensus and Its Application To Inpainting I. R. Paulino and I. Hounie We propose a novel framework for this type of problem based on the idea that estimated patches should coincide at the overlaps (consensus), and develop an algorithm for solving the general problem.
1158 Image Recovery from Rotational And Translational Invariants N. F. Marshall, T. Lan, T. Bendory and A. Singer We introduce a framework for recovering an image from its rotationally and translationally invariant features based on autocorrelation analysis.
1159 Optimal Window Design for Joint Spatial-Spectral Domain Filtering of Signals on the Sphere A. Aslam and Z. Khalid We present the optimal design of an azimuthally symmetric window signal for carrying out joint spatial-spectral domain filtering of a spherical (source) signal contaminated by a realization of an anisotropic noise process.
1160 Filtering Out Time-Frequency Areas Using Gabor Multipliers A. M. Kr?m?, V. Emiya, C. Chaux and B. Torr?sani We address the problem of filtering out localized time-frequency components in signals.
1161 ?-NMF and Sparsity Promoting Regularizations for Complex Mixture Unmixing. Application to 2D HSQC NMR. A. CHERNI, S. ANTHOINE and C. CHAUX In this work, we propose a new variational formulation for blind source separation (BSS) based on a ?-divergence data fidelity term combined with sparsity promoting regularization functions.
1162 Fir Filtering of Discontinuous Signals: A Random-Stratified Sampling Approachx H. Y. Darawsheh and A. Tarczynski This paper presents a novel approach, based on random stratified sampling (StSa) technique, to estimate the output of a finite impulse response (FIR) filter when the input signal is either a piecewise-continuous function having first-derivative discontinuities (FDDs), or a piecewise-discontinuous function, i.e. having zero-derivative discontinuities (ZDDs).
1163 Message Transmission ThroughUnderspread Time-Varying Linear Channels A. Kaplan, D. Lee and V. Pohl The main contribution of this paper is that the suggested message transmission scheme over time-variant communication channels enables data transmission in scenarios where previously no communication was possible.
1164 The Discrete Stockwell Transforms for Infinite-Length Signals and Their Real-Time Implementations Y. Yan and H. Zhu In this paper, new formulations of the discrete ST for infinite-length signals are proposed.
1165 Low-Rank Approximation of Matrices Via A Rank-Revealing Factorization with Randomization M. F. Kaloorazi and J. Chen To treat such a case efficiently, in this paper we present an algorithm called randomized pivoted TSOD (RP-TSOD) that constructs a highly accurate approximation to the TSOD decomposition.
1166 Time-Scale Synthesis for Locally Stationary Signals A. Meynard and B. Torr?sani In a maximum a posteriori approach, we propose an estimator for the model parameters, namely the time-varying scale translation and an underlying power spectrum.
1167 Maximally Energy-Concentrated Differential Window for Phase-Aware Signal Processing Using Instantaneous Frequency T. Kusano, K. Yatabe and Y. Oikawa In this paper, we propose window functions suitable for computing the instantaneous frequency which are designed based on minimizing the sidelobe energy of the frequency response of the differential window.
1168 On the Use of R?nyi Entropy for Optimal Window Size Computation in the Short-Time Fourier Transform S. Meignen, M. Colominas and D. Pham In this paper, we investigate the determination of an optimal window length associated with the computation of the short time Fourier transform of multicomponent signals.
1169 Data-Driven Model Set Design for Model Averaged Particle Filter B. Liu This paper fills in this gap by proposing a generic data-driven method for BMAPF model set design.
1170 Graphem: EM Algorithm for Blind Kalman Filtering Under Graphical Sparsity Constraints ?. Chouzenoux and V. Elvira In this work, we propose a novel expectation-maximization algorithm for estimating the linear matrix operator in the state equation of a linear-Gaussian state-space model.
1171 On Design of Optimal Smart Meter Privacy Control Strategy Against Adversarial Map Detection R. R. Avula and T. J. Oechtering We study the optimal control problem of the maximum a posteriori (MAP) state sequence detection of an adversary using smart meter data.
1172 Approximate Inference by Kullback-Leibler Tensor Belief Propagation P. W. A. Wijnings, S. Stuijk, B. de Vries and H. Corporaal We focus on reducing the size of discrete models with large treewidth by storing intermediate factors in compressed form, thereby decoupling the variables through conditioning on introduced weights.This work proposes pruning of these weights using Kullback-Leibler divergence.
1173 A Particle Gibbs Sampling Approach to Topology Inference in Gene Regulatory Networks M. Iloska, Y. El-Laham and M. F. Bugallo In this paper, we propose a novel Bayesian approach for estimating a gene network?s topology using particle Gibbs sampling.
1174 Particle Filter with Rejection Control and Unbiased Estimator of the Marginal Likelihood J. Kudlicka, L. M. Murray, T. B. Sch?n and F. Lindsten In the paper we present a particle filter with rejection control that enables unbiased estimation of the marginal likelihood.
1175 Particle Group Metropolis Methods for Tracking the Leaf Area Index L. Martino, V. Elvira and G. Camps-Valls In this work, we introduce an Markov Chain Monte Carlo (MCMC) technique driven by a particle filter.
1176 Unsupervised Variational Bayesian Kalman Filtering For Large-Dimensional Gaussian Systems B. Ait-El-Fquih, T. Rodet and I. Hoteit For this problem, we propose two efficient algorithms based on the variational Bayesian (VB) approach.
1177 Levenberg-Marquardt and Line-Search Extended Kalman Smoothers S. S?rkk? and L. Svensson The aim of this article is to present Levenberg-Marquardt and line-search extensions of the classical iterated extended Kalman smoother (IEKS) which has previously been shown to be equivalent to the Gauss-Newton method.
1178 Laplace State Space Filter with Exact Inference and Moment Matching J. Neri, P. Depalle and R. Badea We present a Bayesian filter for state space models with Laplace-distributed observation noise that is robust to heavy-tailed and outlier-ridden univariate time-series data.
1179 Probabilistic Filter and Smoother for Variational Inference of Bayesian Linear Dynamical Systems J. Neri, R. Badeau and P. Depalle Here, we propose a solution using matrix inversion lemmas to derive what may be considered as the Bayesian counterparts to the Kalman filter and smoother, which are particular forms of the forward-backward algorithm that have known properties of numerical stability and efficiency that lead to cost growing linear with time.
1180 Optimum Kernel Particle Filter for Asymmetric Laplace Noise U. Andersson and S. Godsill In this paper we present on-line Bayesian filtering methods for time series models corrupted by asymmetric Laplace noise.
1181 Convex Optimisation-Based Privacy-Preserving Distributed Average Consensus in Wireless Sensor Networks Q. Li, R. Heusdens and M. G. Christensen In this paper, we propose a novel convex optimization-based solution to the problem of privacy-preserving distributed average consensus.
1182 Proximal Multitask Learning Over Distributed Networks with Jointly Sparse Structure D. Jin, J. Chen, C. Richard and J. Chen By introducing an l8,1-norm penalty at each node, and using a proximal gradient method to minimize the regularized cost, we devise a proximal multitask diffusion LMS algorithm which promotes the joint-sparsity to enhance the estimation performance.
1183 Optimal Sampling Rate and Bandwidth of Bandlimited Signals?an Algorithmic Perspective H. Boche and U. J. M?nich In this paper we study if it is possible to algorithmically determine the actual bandwidth of a bandlimited signal.
1184 Resilient to Byzantine Attacks Finite-Sum Optimization Over Networks Z. Wu, Q. Ling, T. Chen and G. B. Giannakis This contribution deals with distributed finite-sum optimization for learning over networks in the presence of malicious Byzantine attacks.
1185 Exploiting Sparsity for Robust Sensor Network Localization in Mixed LOS/NLOS Environments D. Jin, F. Yin, M. Fau?, M. Muma and A. M. Zoubir We address the problem of robust network localization in realistic mixed LOS/NLOS environments.
1186 A Low-Complexity Map Detector for Distributed Networks A. E. Feitosa, V. H. Nascimento and C. G. Lopes This work describes a generalization of our previous maximum likelihood (ML) detector to a maximum a posteriori (MAP) detector in distributed networks using the diffusion LMS algorithm.
1187 Quickest Change Detection In Anonymous Heterogeneous Sensor Networks Z. Sun, S. Zou and Q. Li In this paper, a simple optimality proof is derived for the Mixture Likelihood Ratio Test (MLRT), which was constructed and proved to be optimal for the non-sequential anonymous setting in [1].
1188 Optimal Power Flow Using Graph Neural Networks D. Owerko, F. Gama and A. Ribeiro In this paper, we propose using graph neural networks (which are localized, scalable parametrizations of network data) trained under the imitation learning framework to approximate a given optimal solution.
1189 Byzantine-Robust Decentralized Stochastic Optimization J. Peng and Q. Ling In this paper, we consider the Byzantine-robust stochastic optimization problem defined over a decentralized network, where the agents collaboratively minimize the summation of expectations of stochastic local cost functions, but some of the agents are unreliable.
1190 Federated Truth Inference over Distributed Crowdsourcing Platforms M. Yang, G. Liu and Y. -. P. Hong A federated truth inference (FTI) algorithm is proposed based on a distributed implementation of the block expectation-maximization (EM) algorithm.
1191 Clock Synchronization Over Networks Using Sawtooth Models P. d. Aguila Pla, L. Pellaco, S. Dwivedi, P. H?ndel and J. Jald?n In this paper, we study the use of time-to-digital converters in wireless sensors, which provides clock synchronization and ranging at negligible communication overhead through a sawtooth signal model for round trip times between two nodes.
1192 An Analytical Solution to Jacobsen Estimator for Windowed Signals T. Murakami and W. Wang In this paper, we extend the well-known Jacobsen estimator to windowed signals.
1193 Regularized Partial Phase Synchrony Index Applied to Dynamical Functional Connectivity Estimation G. Frusque, J. Jung, P. Borgnat and P. Gon?alves We study the inference of conditional independence graph from the partial Phase Locking Value (PLV) index of multivariate time series.
1194 The Matched Reassigned Cross-Spectrogram for Phase Estimation M. Sandsten, R. Anderson, I. Reinhold and J. Brynolfsson In this paper, the matched reassigned spectrogram is expanded into a novel matched phase reassignment (MPR) method based on the reassigned cross-spectrogram.
1195 Line Spectral Estimation with Palindromic Kernels D. Verbeke and I. Markovsky We develop an equivalent formulation as a structured low-rank approximation problem and present a necessary condition for the model to be undamped.
1196 Latent Fused Lasso Y. Feng and I. Selesnick In this paper, we propose a convex variational norm for better modeling sparse piecewise constant signals.
1197 Adversarial Attacks on Deep Unfolded Networks for Sparse Coding Y. Wang, K. Wu and C. Zhang Our paper is the first work to study the adversarial performance on unfolded DNNs for sparse coding.
1198 Riemannian Framework for Robust Covariance Matrix Estimation in Spiked Models F. Bouchard, A. Breloy, G. Ginolhac and F. Pascal This paper aims at providing an original Riemannian geometry to derive robust covariance matrix estimators in spiked models (i.e. when the covariance matrix has a low-rank plus identity structure).
1199 Robust Matrix Completion via lP-Greedy Pursuits X. Jiang, A. M. Zoubir and X. Liu A novel l p -greedy pursuit (GP) algorithm for robust matrix completion, i.e., recovering a low-rank matrix from only a subset of its noisy and outlier-contaminated entries, is devised.
1200 Separable Optimization for Joint Blind Deconvolution and Demixing D. Weitzner and R. Giryes In this work, we present a separable approach to blind deconvolution and demixing via convex optimization.
1201 Misspecified Cramer-Rao Bound For Delay Estimation with a Mismatched Waveform: A Case Study F. Roemer In this paper we investigate the problem of time of arrival estimation which occurs in many real-world applications, such as indoor localization or non-destructive testing via ultrasound or radar.
1202 Privacy-Aware Quickest Change Detection T. S. Lau and W. Peng Tay Our goal is to sanitize the signal to satisfy information privacy requirements while being able to detect a change quickly.
1203 Source Enumeration via Toeplitz Matrix Completion V. Garg, P. Gim?nez-Febrer, A. Pag?s-Zamora and I. Santamaria This paper addresses the problem of source enumeration by an array of sensors in the presence of noise whose spatial covariance structure is a diagonal matrix with possibly different variances, referred to non-iid noise hereafter, when the sources are uncorrelated.
1204 Sequential Methods for Detecting a Change in the Distribution of an Episodic Process T. Banerjee, E. Adib, A. Taha and E. John The algorithms can be computed recursively using finite memory and are shown to be asymptotically optimal for well-defined Bayesian or minimax stochastic optimization formulations.
1205 Distribution of the Product of a Complex Gaussian Matrix and Vector and Its Sum with a Complex Gaussian Vector W. Shi, Y. Li and Q. He In this paper, we derive the distribution of the product of a complex Gaussian matrix and a complex Gaussian vector.
1206 Principal Angle Detector for Subspace Signal with Structured Unknown Interference X. Xu, Y. Jiao and Y. Gu In this paper, we propose a new detector, called Principal Angle Detector (PAD), based on principal angles between subspaces.
1207 A Robust Speaker Clustering Method Based on Discrete Tied Variational Autoencoder C. Feng, J. Wang, T. Li, J. Peng and J. Xiao In this paper, we propose a novel speaker clustering method based on Mutual Information (MI) and a non-linear model with discrete variable, which under the enlightenment of Tied Variational Autoencoder (TVAE), to enhance the robustness against noise.
1208 Track-Before-Detect for Sub-Nyquist Radar S. Na, T. Huang, Y. Liu and X. Wang To overcome this issue, we propose a weighted sparse recovery based track-before-detect (TBD) method for weak targets detection by accumulating multi-frame information.
1209 Optimal Transport Based Change Point Detection and Time Series Segment Clustering K. C. Cheng, S. Aeron, M. C. Hughes, E. Hussey and E. L. Miller Building upon recent theoretical advances characterizing the limiting distribution-free behavior of the Wasserstein two-sample test (Ramdas et al. 2015), we propose a novel algorithm for unsupervised, distribution-free CPD which is amenable to both offline and online settings.
1210 Multi-View Wasserstein Discriminant Analysis with Entropic Regularized Wasserstein Distance H. Kasai To evaluate this discrepancy, this paper presents a proposal of a multi-view Wasserstein discriminant analysis, designated as MvWDA, which exploits a recently developed optimal transport theory.
1211 Large-Scale Time Series Clustering with k-ARs Z. Yue and V. Solo This paper proposes a new algorithm, k-ARs, which is a limiting version of the existing EM algorithm.
1212 Deterministic Feature Decoupling by Surfing Invariance Manifolds E. Mart?nez-Enr?quez and J. Portilla We introduce a formalism that justifies and extends a heuristic method for algebraically decoupling deterministic features that recently proved useful for improving feature-based classifi-cation.
1213 Unsupervised Auto-Encoding Multiple-Object Tracker for Constraint-Consistent Combinatorial Problem Y. Kawachi and T. Suzuki In this work, we propose an unsupervised neural MOT model for accurate semi-automatic association labeling and we tackle the challenging one-to-one constrained combinatorial association problem by applying relaxation techniques.
1214 A Streaming On-Device End-To-End Model Surpassing Server-Side Conventional Model Quality and Latency T. N. Sainath et al. In this paper, we develop a first-pass Recurrent Neural Network Transducer (RNN-T) model and a second-pass Listen, Attend, Spell (LAS) rescorer that surpasses a conventional model in both quality and latency.
1215 Minimum Latency Training Strategies for Streaming Sequence-to-Sequence ASR H. Inaguma, Y. Gaur, L. Lu, J. Li and Y. Gong To alleviate this issue and reduce latency, we propose several strategies during training by leveraging external hard alignments extracted from the hybrid model.
1216 Towards Fast and Accurate Streaming End-To-End ASR B. Li et al. We propose to reduce E2E model?s latency by extending the RNN-T endpointer (RNN-T EP) model [2] with additional early and late penalties.
1217 Streaming Automatic Speech Recognition with the Transformer Model N. Moritz, T. Hori and J. Le In this work, we propose a transformer based end-to-end ASR system for streaming ASR, where an output must be generated shortly after each spoken word.
1218 CIF: Continuous Integrate-And-Fire for End-To-End Speech Recognition L. Dong and B. Xu In this paper, we propose a novel soft and monotonic alignment mechanism used for sequence transduction.
1219 Transformer-Based Online CTC/Attention End-To-End Speech Recognition Architecture H. Miao, G. Cheng, C. Gao, P. Zhang and Y. Yan In this paper, we propose the Transformer-based online CTC/attention E2E ASR architecture, which contains the chunk self-attention encoder (chunk-SAE) and the monotonic truncated attention (MTA) based self-attention decoder (SAD).
1220 Detecting Multiple Speech Disfluencies Using a Deep Residual Network with Bidirectional Long Short-Term Memory T. Kourkounakis, A. Hajavi and A. Etemad As opposed to most existing works that identify stutters with language models, our work proposes a model that relies solely on acoustic features, allowing for identification of several variations of stutter disfluencies without the need for speech recognition.
1221 Exploring Appropriate Acoustic and Language Modelling Choices for Continuous Dysarthric Speech Recognition Z. Yue, F. Xiong, H. Christensen and J. Barker In this paper, we investigate the impact of LM design using the widely used TORGO database.
1222 Synthetic Speech References for Automatic Pathological Speech Intelligibility Assessment P. Janbakhshi, I. Kodrasi and H. Bourlard To be able to use P-ESTOI in such scenarios, in this paper we propose to use synthetic speech generated by state-of-the-art high-quality text-to-speech systems to create an intelligible reference model.
1223 Two-Step Acoustic Model Adaptation for Dysarthric Speech Recognition R. Takashima, T. Takiguchi and Y. Ariki This paper introduces a model adaptation approach for a speaker-dependent dysarthric speech recognition system.
1224 Dysarthric Speech Recognition with Lattice-Free MMI E. Hermann and M. Magimai.-Doss This paper focuses on the use of state-of-the-art sequence-discriminative training, in particular lattice-free maximum mutual information (LF-MMI), for improving dysarthric speech recognition.
1225 Improved Speaker Independent Dysarthria Intelligibility Classification Using Deepspeech Posteriors A. Tripathi, S. Bhosale and S. K. Kopparapu Motivated by this observation, we propose a speaker independent intelligibility assessment system which relies on a novel set of features obtained by processing the output of DeepSpeech, an end to end Speech-to-Text engine.
1226 Joint Phoneme-Grapheme Model for End-To-End Speech Recognition Y. Kubo and M. Bacchiani This paper proposes methods to improve a commonly used end-to-end speech recognition model, Listen-Attend-Spell (LAS).
1227 Quartznet: Deep Automatic Speech Recognition with 1D Time-Channel Separable Convolutions S. Kriman et al. We propose a new end-to-end neural acoustic model for automatic speech recognition.
1228 End-To-End Multi-Talker Overlapping Speech Recognition A. Tripathi, H. Lu and H. Sak In this paper we present an end-to-end speech recognition system that can recognize single-channel speech where multiple talkers can speak at the same time (overlapping speech) by using a neural network model based on Recurrent Neural Network Transducer (RNN-T) architecture.
1229 End-To-End Multi-Speaker Speech Recognition With Transformer X. Chang, W. Zhang, Y. Qian, J. L. Roux and S. Watanabe In this work, we explore the use of Transformer models for these tasks by focusing on two aspects.
1230 Hybrid Autoregressive Transducer (HAT) E. Variani, D. Rybach, C. Allauzen and M. Riley This paper proposes and evaluates the hybrid autoregressive transducer (HAT) model, a time-synchronous encoder-decoder model that preserves the modularity of conventional automatic speech recognition systems.
1231 Lightweight and Efficient End-To-End Speech Recognition Using Low-Rank Transformer G. I. Winata, S. Cahyawijaya, Z. Lin, Z. Liu and P. Fung We propose the low-rank transformer (LRT), a memory-efficient and fast neural architecture that significantly reduces the parameters and boosts the speed of training and inference for end-to-end speech recognition.
1232 Spoken Language Acquisition Based on Reinforcement Learning and Word Unit Segmentation S. Gao, W. Hou, T. Tanaka and T. Shinozaki This paper proposes a new framework for simulating spoken-language acquisition by combining reinforcement learning and unsupervised learning methods.
1233 How Much Self-Attention Do We Need? Trading Attention for Feed-Forward Layers K. Irie, A. Gerstenberger, R. Schl?ter and H. Ney We propose simple architectural modifications in the standard Transformer with the goal to reduce its total state size (defined as the number of self-attention layers times the sum of the key and value dimensions, times position) without loss of performance.
1234 Learning Recurrent Neural Network Language Models With Context-Sensitive Label Smoothing for Automatic Speech Recognition M. Song, Y. Zhao, S. Wang and M. Han We propose an approach of context-sensitive candidate label smoothing that has two advantages.
1235 Semi-Supervised Learning for Text Classification by Layer Partitioning A. H. Li and A. Sethy To adapt these methods to text input, we propose to decompose a neural network M into two components F and U so that M = U ?
1236 Integrating Discrete and Neural Features Via Mixed-Feature Trans-Dimensional Random Field Language Models S. Gao, Z. Ou, W. Yang and H. Xu This paper develops a mixed-feature TRF LM and demonstrates its advantage in integrating discrete and neural features.
1237 Gated Attentive Convolutional Network Dialogue State Tracker S. Liu, S. Liu and W. Xu In this paper, we propose the Gated Attentive Convolutional network Dialogue State Tracker (GAC) which overcomes these challenges by utilizing the gated attentive convolutional encoder and introducing historical information.
1238 Using Vaes and Normalizing Flows for One-Shot Text-To-Speech Synthesis of Expressive Speech V. Aggarwal, M. Cotescu, N. Prateek, J. Lorenzo-Trueba and R. Barra-Chicote We propose a Text-to-Speech method to create an unseen expressive style using one utterance of expressive speech of around one second.
1239 Zero-Shot Multi-Speaker Text-To-Speech with State-Of-The-Art Neural Speaker Embeddings E. Cooper et al. We investigate multi-speaker modeling for end-to-end text-to-speech synthesis and study the effects of different types of state-of-the-art neural speaker embeddings on speaker similarity for unseen speakers.
1240 Mellotron: Multispeaker Expressive Voice Synthesis by Conditioning on Rhythm, Pitch and Global Style Tokens R. Valle, J. Li, R. Prenger and B. Catanzaro Mellotron is a multispeaker voice synthesis model based on Tacotron 2 GST that can make a voice emote and sing without emotive or singing training data.
1241 Location-Relative Attention Mechanisms for Robust Long-Form Speech Synthesis E. Battenberg et al. We suggest simple modifications to GMM-based attention that allow it to align quickly and consistently during training, and introduce a new location-relative attention mechanism to the additive energy-based family, called Dynamic Convolution Attention (DCA).
1242 Parallel Wavegan: A Fast Waveform Generation Model Based on Generative Adversarial Networks with Multi-Resolution Spectrogram R. Yamamoto, E. Song and J. Kim We propose Parallel WaveGAN, a distillation-free, fast, and small-footprint waveform generation method using a generative adversarial network.
1243 Gaussian Lpcnet for Multisample Speech Synthesis V. Popov, M. Kudinov and T. Sadekova In this work, we present a modification of LPCNet that is 1.5x faster, has twice less non-zero parameters and synthesizes speech of the same quality.
1244 A Computationally Light Algorithm for Bayesian Speech Enhancement with SNR Marginalization S. Thaleiser and G. Enzner Specifically, the extrema of the posterior distribution, which can easily be obtained via differentiation, are combined according to their widths, heights and abscissa.
1245 Low-Latency Single Channel Speech Enhancement Using U-Net Convolutional Neural Networks A. E. Bulut and K. Koishida To do this, we propose a simple but effective U-Net convolutional neural network (CNN) based architecture with skip-connections with a focus on real-time applications which require low-latency processing.
1246 A Cross-Task Transfer Learning Approach to Adapting Deep Speech Enhancement Models to Unseen Background Noise Using Paired Senone Classifiers S. Wang, W. Li, S. M. Siniscalchi and C. Lee We propose an environment adaptation approach that improves deep speech enhancement models via minimizing the Kullback-Leibler divergence between posterior probabilities produced by a multi-condition senone classifier (teacher) fed with noisy speech features and a clean-condition senone classifier (student) fed with enhanced speech features to transfer an existing deep neural network (DNN) speech enhancer to specific noisy environments without using noisy/clean paired target waveforms needed in conventional DNN-based spectral regression.
1247 Monaural Speech Enhancement Using Intra-Spectral Recurrent Layers in the Magnitude and Phase Responses K. M. Nayem and D. S. Williamson In this paper, we propose a deep learning architecture that leverages both temporal and spectral dependencies within the magnitude and phase responses.
1248 A Maximum Likelihood Approach to Multi-Objective Learning Using Generalized Gaussian Distributions for Dnn-Based Speech Enhancement S. Niu, J. Du, L. Chai and C. Lee In this paper, we extend the maximum likelihood approach proposed in our previous work [1] to the multi-objective learning for DNN-based speech enhancement (ML-MOL-DNN) to achieve the automatic adjustment of the dynamic range of prediction error values on different targets.
1249 PAGAN: A Phase-Adapted Generative Adversarial Networks for Speech Enhancement P. Li et al. This paper presents a new approach to solve the phase mismatch problem by training traditional DNN adversarially with a time-domain discriminator.
1250 Humangan: Generative Adversarial Network With Human-Based Discriminator And Its Evaluation In Speech Perception Modeling K. Fujii, Y. Saito, S. Takamichi, Y. Baba and H. Saruwatari We propose the HumanGAN, a generative adversarial network (GAN) incorporating human perception as a discriminator.
1251 The Processing of Mandarin Chinese Tonal Alternations in Contexts: An Eye-Tracking Study J. Tu and Y. Chien This study investigated the perception of Mandarin tonal alternations in disyllabic words.
1252 On The Impact of Language Familiarity in Talker Change Detection N. Sharma, V. Krishnamohan, S. Ganapathy, A. Gangopadhayay and L. Fink In this paper, we propose an experimental paradigm to provide insights on the impact of language familiarity on talker change detection.
1253 Effects of Spectral Tilt on Listeners? Preferences And Intelligibility O. Simantiraki, M. Cooke and Y. Pantazis This paper describes a real-time speech modification technique which allows evaluation of the impact of individual speech properties on listeners? preferences.
1254 Effect of Frication Duration and Formant Transitions on the Perception of Fricatives in VCV Utterances K. S. Nataraj, P. C. Pandey and H. Dasgupta A study is conducted to investigate the relative importance of the transitions and the frication duration on the perception of the unvoiced fricatives /f, s, sh/.
1255 Fully-Hierarchical Fine-Grained Prosody Modeling For Interpretable Speech Synthesis G. Sun, Y. Zhang, R. J. Weiss, Y. Cao, H. Zen and Y. Wu This paper proposes a hierarchical, fine-grained and interpretable latent variable model for prosody based on the Tacotron 2 text-to-speech model.
1256 Transferring Neural Speech Waveform Synthesizers to Musical Instrument Sounds Generation Y. Zhao, X. Wang, L. Juvela and J. Yamagishi This work compares three neural synthesizers used for musical instrument sounds generation under three scenarios: training from scratch on music data, zero-shot learning from the speech domain, and fine-tuning-based adaptation from the speech to the music domain.
1257 Teacher-Student Training For Robust Tacotron-Based TTS R. Liu, B. Sisman, J. Li, F. Bao, G. Gao and H. Li To overcome this, we propose a teacher-student training scheme for Tacotron-based TTS by introducing a distillation loss function in addition to the feature loss function.
1258 Many-To-Many Voice Conversion Using Conditional Cycle-Consistent Adversarial Networks S. Lee, B. Ko, K. Lee, I. Yoo and D. Yook In this paper, we extend the CycleGAN by conditioning the network on speakers.
1259 F0-Consistent Many-To-Many Non-Parallel Voice Conversion Via Conditional Autoencoder K. Qian, Z. Jin, M. Hasegawa-Johnson and G. J. Mysore In the paper, we modified and improved autoencoder-based voice conversion to disentangle content, F0, and speaker identity at the same time.
1260 End-To-End Accent Conversion Without Using Native Utterances S. Liu et al. In this paper, we present an end-to-end framework, which is able to conduct AC from non-native-accented utterances without using any native-accented utterances during online conversion.
1261 Cogans For Unsupervised Visual Speech Adaptation To New Speakers A. Fernandez-Lopez, A. Karaali, N. Harte and F. M. Sukno Specifically, we are the first to explore the visual adaptation of an SI-AVSR system to an unknown and unlabelled speaker.
1262 Visually Guided Self Supervised Learning of Speech Representations A. Shukla, K. Vougioukas, P. Ma, S. Petridis and M. Pantic We propose a framework for learning audio representations guided by the visual modality in the context of audiovisual speech.
1263 Looking Enhances Listening: Recovering Missing Speech Using Images T. Srinivasan, R. Sanabria and F. Metze In this paper, we present a set of experiments where we show the utility of the visual modality under noisy conditions.
1264 Towards Multilingual Sign Language Recognition S. Tornay, M. Razavi and M. Magimai.-Doss In this paper, we develop a multilingual sign language approach, where hand movement modeling is also done with target sign language independent data by derivation of hand movement subunits.
1265 Automatic Identification of Speakers From Head Gestures in a Narration S. K. Vadiraj, A. Rao M V and P. K. Ghosh In this work, we focus on quantifying speaker identity information encoded in the head gestures of speakers, while they narrate a story.
1266 Lipreading Using Temporal Convolutional Networks B. Martinez, P. Ma, S. Petridis and M. Pantic In this work, we address the limitations of this model and we propose changes which further improve its performance.
1267 On Modeling ASR Word Confidence W. Jeon, M. Jordan and M. Krishnamoorthy We present a new method for computing ASR word confidences that effectively mitigates the effect of ASR errors for diverse downstream applications, improves the word error rate of the 1-best result, and allows better comparison of scores across different models.
1268 Confidence Estimation for Black Box Automatic Speech Recognition Systems Using Lattice Recurrent Neural Networks A. Kastanos, A. Ragni and M. J. F. Gales This paper examines this limited resource scenario for confidence estimation, a measure commonly used to assess transcription reliability.
1269 OOV Recovery with Efficient 2nd Pass Decoding and Open-vocabulary Word-level RNNLM Rescoring for Hybrid ASR X. Zhang, D. Povey and S. Khudanpur In this paper, we investigate out-of-vocabulary (OOV) word recovery in hybrid automatic speech recognition (ASR) systems, with emphasis on dynamic vocabulary expansion for both Weight Finite State Transducer (WFST)-based decoding and word-level RNNLM rescoring.
1270 End to End Speech Recognition Error Prediction with Sequence to Sequence Learning P. Serai, A. Stiff and E. Fosler-Lussier We present a novel end to end model to simulate ASR errors.
1271 ASR Error Correction and Domain Adaptation Using Machine Translation A. Mani, S. Palaskar, N. V. Meripo, S. Konam and F. Metze We propose a simple technique to perform domain adaptation for ASR error correction via machine translation.
1272 Joint Contextual Modeling for ASR Correction and Language Understanding Y. Weng et al. In this paper, we propose multi-task neural approaches to perform contextual language correction on ASR outputs jointly with LU to improve the performance of both tasks simultaneously.
1273 Deep Casa for Talker-independent Monaural Speech Separation Y. Liu, M. Delfarah and D. Wang In this study, we address both speech and nonspeech interference, i.e., monaural speaker separation in noise, in a talker-independent fashion.
1274 Demystifying TasNet: A Dissecting Approach J. Heitkaemper, D. Jakobeit, C. Boeddeker, L. Drude and R. Haeb-Umbach In this paper we dissect the gains of the time-domain audio separation network (TasNet) approach by gradually replacing components of an utterance-level permutation invariant training (u-PIT) based separation system in the frequency domain until the TasNet system is reached, thus blending components of frequency domain approaches with those of time domain approaches.
1275 Filterbank Design for End-to-end Speech Separation M. Pariente, S. Cornell, A. Deleforge and E. Vincent In this work, we extend real-valued learned and parameterized filterbanks into complex-valued analytic filterbanks and define a set of corresponding representations and masking strategies.
1276 Interrupted and Cascaded Permutation Invariant Training for Speech Separation G. Yang, S. Wu, Y. Mao, H. Lee and L. Lee In this paper, we investigate instead for a given model architecture the various flexible label assignment strategies for training the model, rather than directly using PIT.
1277 Mixup-breakdown: A Consistency Training Method for Improving Generalization of Speech Separation Models M. W. Y. Lam, J. Wang, D. Su and D. Yu To address this problem, we propose an easy-to-implement yet effective consistency based semi-supervised learning (SSL) approach, namely Mixup-Breakdown training (MBT).
1278 An Online Speaker-aware Speech Separation Approach Based on Time-domain Representation H. Wang, Y. Song, Z. Li, I. McLoughlin and L. Dai In this paper, we propose a speaker-aware approach based on the source-aware context network structure, in which the speaker information is explicitly modeled by an auxiliary speaker identification branch.
1279 Beam-TasNet: Time-domain Audio Separation Network Meets Frequency-domain Beamformer T. Ochiai, M. Delcroix, R. Ikeshita, K. Kinoshita, T. Nakatani and S. Araki Motivated by this question, this paper proposes a novel speech separation scheme, i.e., Beam-TasNet, which combines TasNet with the frequency-domain beamformer, i.e., a minimum variance distortionless response (MVDR) beamformer, through spatial covariance computation to achieve better ASR performance.
1280 On End-to-end Multi-channel Time Domain Speech Separation in Reverberant Environments J. Zhang, C. Zorila, R. Doddipatla and J. Barker This paper introduces a new method for multi-channel time domain speech separation in reverberant environments.
1281 End-to-end Microphone Permutation and Number Invariant Multi-channel Speech Separation Y. Luo, Z. Chen, N. Mesgarani and T. Yoshioka In this paper, we propose transform-average-concatenate (TAC), a simple design paradigm for channel permutation and number invariant multi-channel speech separation.
1282 DNN-supported Mask-based Convolutional Beamforming for Simultaneous Denoising, Dereverberation, and Source Separation T. Nakatani et al. In this article, we investigate an integrated mask-based convolutional beamforming method for performing simultaneous denoising, dereverberation, and source separation.
1283 Real-Time Binaural Speech Separation with Preserved Spatial Cues C. Han, Y. Luo and N. Mesgarani Here, we propose a speech separation algorithm that preserves the interaural cues of separated sound sources and can be implemented with low latency and high fidelity, therefore enabling a real-time modification of the acoustic scene.
1284 SLOGD: Speaker Location Guided Deflation Approach to Speech Separation S. Sivasankaran, E. Vincent and D. Fohr In this work we propose to separate the sources using a Speaker LOcalization Guided Deflation (SLOGD) approach wherein we estimate the sources iteratively.
1285 Multilingual Acoustic Word Embedding Models for Processing Zero-resource Languages H. Kamper, Y. Matusevych and S. Goldwater Here we propose to train a single supervised embedding model on labelled data from multiple well-resourced languages and then apply it to unseen zero-resource languages.
1286 Mockingjay: Unsupervised Speech Representation Learning with Deep Bidirectional Transformer Encoders A. T. Liu, S. Yang, P. Chi, P. Hsu and H. Lee We present Mockingjay as a new speech representation learning approach, where bidirectional Transformer encoders are pre-trained on a large amount of unlabeled speech.
1287 Recurrent Neural Audiovisual Word Embeddings for Synchronized Speech and Real-Time Mri ?. D. K?se and M. Sara?lar In this paper, the use of word embeddings for the segments found in audio and real-time magnetic resonance imaging (rtMRI) videos is addressed.
1288 Deep Contextualized Acoustic Representations for Semi-Supervised Speech Recognition S. Ling, Y. Liu, J. Salazar and K. Kirchhoff We propose a novel approach to semi-supervised automatic speech recognition (ASR).
1289 What Does a Network Layer Hear? Analyzing Hidden Representations of End-to-End ASR Through Speech Synthesis C. Li, P. Yuan and H. Lee In this paper, we present our ASR probing model, which synthesizes speech from hidden representations of end-to-end ASR to examine the information maintained after each layer calculation.
1290 Learning a Subword Inventory Jointly with End-to-End Automatic Speech Recognition J. Drexler and J. Glass In this paper, we follow the LSD method for using subword units but introduce an updated loss function that allows the ASR model to explicitly perform unit discovery, as well.
1291 Multiple Points Input For Convolutional Neural Networks in Replay Attack Detection S. Yoon and H. Yu We proposed the multiple points input method to increase the amount of information that can be considered at one time.
1292 Information Maximized Variational Domain Adversarial Learning for Speaker Verification Y. Tu, M. Mak and J. Chien This paper proposes an information-maximized variational domain adversarial neural network (InfoVDANN) to reduce domain mismatch by incorporating an InfoVAE into domain adversarial training (DAT).
1293 Text Adaptation for Speaker Verification with Speaker-Text Factorized Embeddings Y. Yang, S. Wang, X. Gong, Y. Qian and K. Yu In this paper, we propose a novel text adaptation framework to address the text mismatch issue.
1294 Voiceai Systems to NIST Sre19 Evaluation: Robust Speaker Recognition on Conversational Telephone Speech R. Li, D. Chen and W. Zhang In this study, we present the VoiceAI (VAI) submissions to NIST SRE 2019 challenge on the task of speaker recognition using conversational telephone speech.
1295 Multi-Resolution Multi-Head Attention in Deep Speaker Embedding Z. Wang, K. Yao, X. Li and S. Fang This paper proposes simple but effective pooling methods to compute attentive weights for better temporal aggregation over the variable-length input speech, enabling the end-to-end neural network to have improved performance for discriminating among speakers.
1296 Within-Sample Variability-Invariant Loss for Robust Speaker Recognition Under Noisy Environments D. Cai, W. Cai and M. Li In this paper, we train the speaker embedding network to learn the “clean” embedding of the noisy utterance.
1297 Speech Emotion Recognition with Dual-Sequence LSTM Architecture J. Wang, M. Xue, R. Culhane, E. Diao, J. Ding and V. Tarokh In this work, we propose a new dual-level model that predicts emotions based on both MFCC features and mel-spectrograms produced from raw audio signals.
1298 A Dialogical Emotion Decoder for Speech Motion Recognition in Spoken Dialog S. Yeh, Y. Lin and C. Lee In this paper, we proposed a novel inference algorithm, a dialogical emotion decoding (DED) algorithm, that treats a dialog as a sequence and consecutively decode the emotion states of each utterance over time with a given recognition engine.
1299 Fusion Approaches for Emotion Recognition from Speech Using Acoustic and Text-Based Features L. Pepino, P. Riera, L. Ferrer and A. Gravano In this paper, we study different approaches for classifying emotions from speech using acoustic and text-based features.
1300 Multi-Time-Scale Convolution for Emotion Recognition from Speech Audio Signals E. Guizzo, T. Weyde and J. B. Leveson To address this and potentially other tasks, we introduce the multi-time-scale (MTS) method to create flexibility towards temporal variations when analyzing time-frequency representations of audio data.
1301 Ordinal Learning for Emotion Recognition in Customer Service Calls W. Han, T. Jiang, Y. Li, B. Schuller and H. Ruan To employ the ordinal information between emotional ranks, we propose to model the ordinal SER tasks under a COnsistent RAnk Logits (CORAL) based deep learning framework.
1302 HGFM : A Hierarchical Grained and Feature Model for Acoustic Emotion Recognition Y. Xu, H. Xu and J. Zou To solve the problem of poor classification performance of multiple complex emotions in acoustic modalities, we propose a hierarchical grained and feature model (HGFM).
1303 Speaker Diarization Using Latent Space Clustering in Generative Adversarial Network M. Pal et al. In this work, we propose deep latent space clustering for speaker diarization using generative adversarial network (GAN) back-projection with the help of an encoder network.
1304 Multimodal Speaker Diarization of Real-World Meetings Using D-Vectors With Spatial Features W. Kang, B. C. Roy and W. Chow In this paper, we present a novel approach to multimodal speaker diarization that combines d-vectors with spatial information derived from performing beamforming given a multi-channel microphone array.
1305 Speaker Diarization with Region Proposal Network Z. Huang et al. In this paper, we propose a novel speaker diarization method: Region Proposal Network based Speaker Diarization (RPNSD).
1306 Optimizing Bayesian Hmm Based X-Vector Clustering for the Second Dihard Speech Diarization Challenge M. Diez, L. Burget, F. Landini, S. Wang and H. Cernock? In this paper, we focus on the two x-vector clustering methods employed, namely Agglomerative Hierarchical Clustering followed by a clustering based on Bayesian Hidden Markov Model (BHMM).
1307 A Memory Augmented Architecture for Continuous Speaker Identification in Meetings N. Flemotomos and D. Dimitriadis We introduce and analyze a novel approach to the problem of speaker identification in multi-party recorded meetings.
1308 But System for the Second Dihard Speech Diarization Challenge F. Landini et al. This paper describes the winning systems developed by the BUT team for the four tracks of the Second DIHARD Speech Diarization Challenge.
1309 Estimating the Degree of Sleepiness by Integrating Articulatory Feature Knowledge in Raw Waveform Based CNNS J. Fritsch, S. P. Dubagunta and M. Magimai.-Doss This paper investigates an end-to-end approach, where given raw waveform as input, a convolutional neural network (CNN) estimates at its output the degree of sleepiness.
1310 Automatic Prediction of Suicidal Risk in Military Couples Using Multimodal Interaction Cues from Couples Conversations S. N. Chakravarthula et al. In this work, we investigate whether acoustic, lexical, behavior and turn-taking cues from military couples? conversations can provide meaningful markers of suicidal risk.
1311 Comparison of User Models Based on GMM-UBM and I-Vectors for Speech, Handwriting, and Gait Assessment of Parkinson?s Disease Patients J. C. Vasquez-Correa, T. Bocklet, J. R. Orozco-Arroyave and E. N?th This study introduces the use of GMM-UBM and i-vectors to evaluate the neurological state of Parkinson?s patients using information from speech, handwriting, and gait.
1312 Exploiting Vocal Tract Coordination Using Dilated CNNS For Depression Detection In Naturalistic Environments Z. Huang, J. Epps and D. Joachim Motivated by the success of these methods, this paper proposes a novel way to extract full vocal tract coordination (FVTC) features by use of convolutional neural networks (CNNs), overcoming earlier shortcomings.
1313 Deep Learning Based Prediction of Hypernasality for Clinical Applications V. C. Mathad, K. Chapman, J. Liss, N. Scherer and V. Berisha In this paper, we propose a new approach that uses the speech samples from healthy controls to model the acoustic characteristics of nasalized speech.
1314 Language Independent Gender Identification from Raw Waveform Using Multi-Scale Convolutional Neural Networks K. D N, A. D, S. S. Reddy, A. Acharya, P. A. Garapati and T. B J In this work, we propose a raw waveform based multiscale convolution neural network approach for language-independent gender identification.
1315 Defense Against Adversarial Attacks on Spoofing Countermeasures of ASV H. Wu, S. Liu, H. Meng and H. Lee In this work, we introduce a passive defense method, spatial smoothing, and a proactive defense method, adversarial training, to mitigate the vulnerability of ASV spoofing countermeasure models against adversarial examples.
1316 Text-Independent Speaker Verification with Adversarial Learning on Short Utterances K. Liu and H. Zhou To address the problem, in this paper, we propose an adversarially learned embedding mapping model that directly maps a short embedding to an enhanced embedding with increased discriminability.
1317 Channel Invariant Speaker Embedding Learning with Joint Multi-Task and Adversarial Training Z. Chen, S. Wang, Y. Qian and K. Yu Inspired by the successful joint multi-task and adversarial training with phonetic information for phonetic-invariant speaker embedding learning, in this paper, a similar methodology is developed to suppress the channel variability.
1318 Adversarial Attacks on GMM I-Vector Based Speaker Verification Systems X. Li, J. Zhong, X. Wu, J. Yu, X. Liu and H. Meng This work investigates the vulnerability of Gaussian Mixture Model (GMM) i-vector based speaker verification systems to adversarial attacks, and the transferability of adversarial samples crafted from GMM i-vector based systems to x-vector based systems.
1319 Orthogonal Training for Text-Independent Speaker Verification Y. Zhu and B. Mak In this paper we propose orthogonal training schemes to improve the effectiveness of cosine similarity measurements in text-independent speaker verification (SV) tasks.
1320 Assessing the Scope of Generalized Countermeasures for Anti-Spoofing R. K. Das, J. Yang and H. Li In this work, we consider widely popular constant-Q cepstral coefficient features along with two other promising front-ends that capture long-term signal characteristics to assess their scope as generalized countermeasures.
1321 Improving Speaker-Attribute Estimation by Voting Based on Speaker Cluster Information N. Tawara, H. Kamiyama, S. Kobashikawa and A. Ogawa This paper proposes a general post-processing method for improving speaker-attribute estimation.
1322 An Ensemble Based Approach for Generalized Detection of Spoofing Attacks to Automatic Speaker Recognizers J. Monteiro, J. Alam and T. H. Falk In this work, we introduce an end-to-end ensemble based approach such that two models ? previously shown to perform well on each considered attack strategy ? are trained jointly, while a third model learns how to mix their outputs yielding a single score.
1323 A Discriminative Condition-Aware Backend for Speaker Verification L. Ferrer and M. McLaren We present a scoring approach for speaker verification that mimics the standard PLDA-based backend process used in most current speaker verification systems.
1324 Adversarial Multi-Task Learning for Speaker Normalization in Replay Detection G. Suthokumar, V. Sethu, K. Sriskandaraja and E. Ambikairajah In this paper, we propose a novel speaker normalisation technique that employs adversarial multi-task learning to compensate for this speaker variability.
1325 Robust Speaker Recognition Using Unsupervised Adversarial Invariance R. Peri, M. Pal, A. Jati, K. Somandepalli and S. Narayanan In this paper, we address the problem of speaker recognition in challenging acoustic conditions using a novel method to extract robust speaker-discriminative speech representations.
1326 A Generalized Framework for Domain Adaptation of PLDA in Speaker Recognition Q. Wang, K. Okabe, K. A. Lee and T. Koshinaka This paper proposes a generalized framework for domain adaptation of Probabilistic Linear Discriminant Analysis (PLDA) in speaker recognition.
1327 CP-GAN: Context Pyramid Generative Adversarial Network for Speech Enhancement G. Liu, K. Gong, X. Liang and Z. Chen In this work, we make the first attempt to explore the global and local speech features for coarse-to-fine speech enhancement and introduce a Context Pyramid Generative Adversarial Network (CPGAN), which contains a densely-connected feature pyramid generator and a dynamic context granularity discriminator to better eliminate audio noise hierarchically.
1328 Densely Connected Neural Network with Dilated Convolutions for Real-Time Speech Enhancement in The Time Domain A. Pandey and D. Wang In this work, we propose a fully convolutional neural network for real-time speech enhancement in the time domain.
1329 Pan: Phoneme-Aware Network for Monaural Speech Enhancement Z. Du, M. Lei, J. Han and S. Zhang Inspired by the progress, we propose a phoneme-aware network (PAN) to utilize the noisy PPGs for speech enhancement.
1330 Efficient Trainable Front-Ends for Neural Speech Enhancement J. Casebeer, U. Isik, S. Venkataramani and A. Krishnaswamy We present an efficient, trainable front-end based on the butterfly mechanism to compute the Fast Fourier Transform, and show its accuracy and efficiency benefits for low-compute neural speech enhancement models.
1331 Invertible DNN-Based Nonlinear Time-Frequency Transform for Speech Enhancement D. Takeuchi, K. Yatabe, Y. Koizumi, Y. Oikawa and N. Harada We propose an end-to-end speech enhancement method with trainable time-frequency (T-F) transform based on invertible deep neural network (DNN).
1332 T-GSA: Transformer with Gaussian-Weighted Self-Attention for Speech Enhancement J. Kim, M. El-Khamy and J. Lee In this paper, we propose a Transformer with Gaussian-weighted self-attention (T-GSA), whose attention weights are attenuated according to the distance between target and context symbols.
1333 Redundant Convolutional Network With Attention Mechanism For Monaural Speech Enhancement T. Lan, Y. Lyu, G. Hui, R. Mokhosi, S. Li and Q. Liu In this study, we introduce an attention mechanism into the convolutional encoderdecoder model.
1334 Residual Recurrent Neural Network for Speech Enhancement J. Abdulbaqi, Y. Gu, S. Chen and I. Marsic We introduce an end-to-end fully recurrent neural network for single-channel speech enhancement.
1335 2D-to-2D Mask Estimation for Speech Enhancement Based on Fully Convolutional Neural Network Y. Tu, J. Du and C. Lee In this study, we design a new Fully CNN (FCNN)-based regression model, which can directly achieve the 2-dimensional (2D) noisy lpg-power spectra (LPS) input to 2dimensional (2D) time-frequency mask output mapping, denoted as 2D-RFCNN.
1336 Self-Supervised Denoising Autoencoder with Linear Regression Decoder for Speech Enhancement R. E. Zezario, T. Hussain, X. Lu, H. Wang and Y. Tsao In this study, we proposed an unsupervised learning approach for speech enhancement, i.e., denoising autoencoder with linear regression decoder (DAELD) model for speech enhancement.
1337 Fully Convolutional Recurrent Networks for Speech Enhancement M. Strake, B. Defraene, K. Fluyt, W. Tirry and T. Fingscheidt As first novelty, we propose to replace the fully-connected LSTM by a convolutional LSTM (ConvLSTM) and call the resulting network a fully convolutional recurrent network (FCRN).
1338 Phonetic Feedback for Speech Enhancement with and Without Parallel Speech Data P. Plantinga, D. Bagchi and E. Fosler-Lussier We use the technique of mimic loss to provide phonetic feedback to an off-the-shelf enhancement system, and find gains in objective intelligibility scores on CHiME-4 data.
1339 Scalable Multilingual Frontend for TTS A. Conkie and A. Finch This paper describes progress towards making a Neural Text-to-Speech (TTS) Frontend that works for many languages and can be easily extended to new languages.
1340 A Unified Sequence-to-Sequence Front-End Model for Mandarin Text-to-Speech Synthesis J. Pan et al. In this paper, we proposed a unified sequence-to-sequence front-end model for Mandarin TTS that converts raw texts to linguistic features directly.
1341 A Hybrid Text Normalization System Using Multi-Head Self-Attention For Mandarin J. Zhang et al. In this paper, we propose a hybrid text normalization system using multi-head self-attention.
1342 Generating Diverse and Natural Text-to-Speech Samples Using a Quantized Fine-Grained VAE and Autoregressive Prosody Prior G. Sun et al. This paper proposes a sequential prior in a discrete latent space which can generate more naturally sounding samples.
1343 Improving Prosody with Linguistic and Bert Derived Features in Multi-Speaker Based Mandarin Chinese Neural TTS Y. Xiao, L. He, H. Ming and F. K. Soong In this study, with a multi-speaker TTS to accommodate the insufficient training data of a target speaker, we investigate linguistic features and Bert-derived information to improve the prosody of our Mandarin Chinese TTS.
1344 Focusing on Attention: Prosody Transfer and Adaptative Optimization Strategy for Multi-Speaker End-to-End Speech Synthesis R. Fu, J. Tao, Z. Wen, J. Yi and T. Wang In this paper, we present two novel methods to handle the above problems by focusing on the attention.
1345 Aligntts: Efficient Feed-Forward Text-to-Speech System Without Explicit Alignment Z. Zeng, J. Wang, N. Cheng, T. Xia and J. Xiao Targeting at both high efficiency and performance, we propose AlignTTS to predict the mel-spectrum in parallel.
1346 GraphTTS: Graph-to-Sequence Modelling in Neural Text-to-Speech A. Sun, J. Wang, N. Cheng, H. Peng, Z. Zeng and J. Xiao This paper leverages the graph-to-sequence method in neural text-to-speech (GraphTTS), which maps the graph embedding of the input sequence to spectrograms.
1347 Effect of Choice of Probability Distribution, Randomness, and Search Methods for Alignment Modeling in Sequence-to-Sequence Text-to-Speech Synthesis Using Hard Alignment Y. Yasuda, X. Wang and J. Yamagishi This research investigates various combinations of sampling methods and probability distributions for alignment transition modeling in a hard-alignment-based sequence-to-sequence TTS method called SSNT-TTS.
1348 Transformer-Based Text-to-Speech with Weighted Forced Attention T. Okamoto, T. Toda, Y. Shiga and H. Kawai Therefore, a Transformer-based acoustic model with weighted forced attention obtained from phoneme durations is proposed to improve synthesis accuracy and stability, where both encoder?decoder attention and forced attention are used with a weighting factor.
1349 Improving End-to-End Speech Synthesis with Local Recurrent Neural Network Enhanced Transformer Y. Zheng, X. Li, F. Xie and L. Lu In this paper, we introduce local recurrent neural network (Local-RNN) into Transformer to make full use of the advantages of both RNN and Transformer while mitigating their drawbacks.
1350 GCI Detection from Raw Speech Using a Fully-Convolutional Network L. Ardaillon and A. Roebel Following this trend, we propose a simple approach that performs a mapping from the speech waveform to a target signal from which the GCIs are obtained by peak-picking.
1351 Frame-Based Overlapping Speech Detection Using Convolutional Neural Networks M. Yousefi and J. H. L. Hansen In this study, we investigate the detection of overlapping speech on segments as short as 25 ms using Convolutional Neural Networks.
1352 Learning Domain Invariant Representations for Child-Adult Classification from Speech R. Lahiri, M. Kumar, S. Bishop and S. Narayanan In this work, we address two major sources of variability-age of the child and data source collection location-using domain adversarial learning which does not require labeled target domain data.
1353 Single Frequency Filter Bank Based Long-Term Average Spectra for Hypernasality Detection and Assessment in Cleft Lip and Palate Speech M. H. Javid, K. Gurugubelli and A. K. Vuppala This work proposes single frequency filter bank based long-term average spectral (SFFB-LTAS) features for hypernasality detection and assessment.
1354 Autoregressive Parameter Estimation with Dnn-Based Pre-Processing Z. Cui, C. Bao, J. K. Nielsen and M. Gr?sb?ll Christensen In this paper, a method for estimating the autoregressive parameters from a signal segment is proposed.
1355 Enhancement of Coded Speech Using a Mask-Based Post-Filter S. Korse, K. Gupta and G. Fuchs In this paper, a data-driven post-filter relying on masking in the time-frequency domain is proposed.
1356 Robust Low Rate Speech Coding Based on Cloned Networks and Wavenet F. S. C. Lim, W. Bastiaan Kleijn, M. Chinen and J. Skoglund We present a new speech-coding scheme that is based on features that are robust to the distortions occurring in speech-coder input signals.
1357 Mixture Factorized Auto-Encoder for Unsupervised Hierarchical Deep Factorization of Speech Signal Z. Peng, S. Feng and T. Lee This paper proposes the mixture factorized auto-encoder (mFAE) for unsupervised deep factorization.
1358 A Novel Approach for Intelligibility Assessment in Dysarthric Subjects A. Tripathi, S. Bhosale and S. K. Kopparapu In the paper, we propose a usable novel method to assess intelligibility of dysarthric speakers.
1359 Voice based classification of patients with Amyotrophic Lateral Sclerosis, Parkinson?s Disease and Healthy Controls with CNN-LSTM using transfer learning J. Mallela et al. In this paper, we consider 2-class and 3-class classification problems for classifying patients with Amyotrophic Lateral Sclerosis (ALS), Parkinson’s Disease (PD), and Healthy Controls (HC) using a CNNLSTM network.
1360 Analysis of Acoustic Features for Speech Sound Based Classification of Asthmatic and Healthy Subjects S. Yadav, M. Keerthana, D. Gope, U. Maheswari K. and P. Kumar Ghosh Analysis of Acoustic Features for Speech Sound Based Classification of Asthmatic and Healthy Subjects
1361 Frequency and Temporal Convolutional Attention for Text-Independent Speaker Recognition S. Yadav and A. Rai In this paper, we propose methods of convolutional attention for independently modelling temporal and frequency information in a convolutional neural network (CNN) based front-end.
1362 Frame-Level Phoneme-Invariant Speaker Embedding for Text-Independent Speaker Recognition on Extremely Short Utterances N. Tawara, A. Ogawa, T. Iwata, M. Delcroix and T. Ogawa This paper investigates a phoneme-invariant speaker embedding approach for speaker recognition on extremely short utterances.
1363 Prototypical Networks for Small Footprint Text-Independent Speaker Verification T. Ko, Y. Chen and Q. Li In this paper, we investigate the use of prototypical networks in a small footprint text-independent speaker verification task.
1364 TDMF: Task-Driven Multilevel Framework for End-to-End Speaker Verification C. Chen and J. Han In this paper, a task-driven multilevel framework (TDMF) is proposed for end-to-end speaker verification.
1365 An Improved Deep Neural Network for Modeling Speaker Characteristics at Different Temporal Scales B. Gu, W. Guo, L. Dai and J. Du This paper presents an improved deep embedding learning method based on a convolutional neural network (CNN) for text-independent speaker verification.
1366 Partial AUC Optimization Based Deep Speaker Embeddings with Class-Center Learning for Text-Independent Speaker Verification Z. Bai, X. Zhang and J. Chen In this paper, we propose a verification loss function, named the maximization of partial area under the Receiver-operating-characteristic (ROC) curve (pAUC), for deep embedding based text-independent speaker verification.
1367 Knowledge Distillation and Random Erasing Data Augmentation for Text-Dependent Speaker Verification V. Mingote, A. Miguel, D. Ribas, A. Ortega and E. Lleida This paper explores the Knowledge Distillation (KD) approach and a data augmentation technique to improve the generalization ability and robustness of text-dependent speaker verification (SV) systems.
1368 Disentangled Speech Embeddings Using Cross-Modal Self-Supervision A. Nagrani, J. S. Chung, S. Albanie and A. Zisserman The objective of this paper is to learn representations of speaker identity without access to manually annotated data.
1369 Improving Deep CNN Networks with Long Temporal Context for Text-Independent Speaker Verification Y. Zhao, T. Zhou, Z. Chen and J. Wu In this paper, we explore two approaches for modeling long temporal contexts to improve the performance of the ResNet networks.
1370 Multi-Level Deep Neural Network Adaptation for Speaker Verification Using MMD and Consistency Regularization W. Lin, M. Mak, N. Li, D. Su and D. Yu In this paper, we present a DNN-based adaptation method using maximum mean discrepancy (MMD).
1371 Multi-Task Learning for Speaker Verification and Voice Trigger Detection S. Sigtia, E. Marchi, S. Kajarekar, D. Naik and J. Bridle In this study, we investigate training a single network to perform both tasks jointly.
1372 Statistics Pooling Time Delay Neural Network Based on X-Vector for Speaker Verification Q. Hong, C. Wu, H. Wang and C. Huang This paper aims to improve speaker embedding representation based on x-vector for extracting more detailed information for speaker verification.
1373 SNDCNN: Self-Normalizing Deep CNNs with Scaled Exponential Linear Units for Speech Recognition Z. Huang, T. Ng, L. Liu, H. Mason, X. Zhuang and D. Liu Inspired by Self-Normalizing Neural Networks, we propose the self-normalizing deep CNN (SNDCNN) based acoustic model topology, by removing the SC/BN and replacing the typical RELU activations with scaled exponential linear unit (SELU) in ResNet-50.
1374 Robust Multi-Channel Speech Recognition Using Frequency Aligned Network T. Park, K. Kumatani, M. Wu and S. Sundaram In this paper, we further develop this idea and use frequency aligned network for robust multi-channel automatic speech recognition (ASR).
1375 Fully Learnable Front-End for Multi-Channel Acoustic Modeling Using Semi-Supervised Learning S. Wager, A. Khare, M. Wu, K. Kumatani and S. Sundaram In this work, we investigated the teacher-student training paradigm to train a fully learnable multi-channel acoustic model for far-field automatic speech recognition (ASR).
1376 G2G: TTS-Driven Pronunciation Learning for Graphemic Hybrid ASR D. Le, T. Koehler, C. Fliegen and M. L. Seltzer In this work, we present a novel method to train a statistical grapheme-to-grapheme (G2G) model on text-to-speech data that can rewrite an arbitrary character sequence into more phonetically consistent forms.
1377 Transformer-Based Acoustic Modeling for Hybrid Speech Recognition Y. Wang et al. We propose and evaluate transformer-based acoustic models (AMs) for hybrid speech recognition.
1378 Specaugment on Large Scale Datasets D. S. Park et al. In this paper, we demonstrate its effectiveness on tasks with large scale datasets by investigating its application to the Google Multidomain Dataset (Narayanan et al., 2018).
1379 Fast Training of Deep Neural Networks for Speech Recognition G. Cong, B. Kingsbury, C. Yang and T. Liu We present an approach for training a bidirectional LSTM acoustic model on the 2000-hour Switchboard corpus.
1380 Unsupervised Pre-Training of Bidirectional Speech Encoders via Masked Reconstruction W. Wang, Q. Tang and K. Livescu We propose an approach for pre-training speech representations via a masked reconstruction loss.
1381 Distilling Attention Weights for CTC-Based ASR Systems T. Moriya, H. Sato, T. Tanaka, T. Ashihara, R. Masumura and Y. Shinohara We present a novel training approach for connectionist temporal classification (CTC) -based automatic speech recognition (ASR) systems.
1382 DEJA-VU: Double Feature Presentation and Iterated Loss in Deep Transformer Networks A. Tjandra et al. Here, we propose to feed the input features at multiple depths in the acoustic model.
1383 Frame-Level MMI as A Sequence Discriminative Training Criterion for LVCSR W. Michel, R. Schl?ter and H. Ney In this work we present frame-level maximum mutual information (MMI) as a novel sequence discriminative training criterion for hybrid HMM-DNN acoustic models.
1384 Cross Lingual Transfer Learning for Zero-Resource Domain Adaptation A. Abad, P. Bell, A. Carmantini and S. Renais We propose a method for zero-resource domain adaptation of DNN acoustic models, for use in low-resource situations where the only in-language training data available may be poorly matched to the intended target domain.
1385 Improving Robustness of Deep Learning Based Monaural Speech Enhancement Against Processing Artifacts K. Tan and D. Wang In this study, we find that enhancing a speech signal twice can dramatically degrade the enhancement performance.
1386 CAD-AEC: Context-Aware Deep Acoustic Echo Cancellation A. Fazel, M. El-Khamy and J. Lee This paper proposes a context- aware deep AEC (CAD-AEC) by introducing two main components.
1387 Artificial Bandwidth Extension Using Conditional Variational Auto-encoders and Adversarial Learning P. Bachhav, M. Todisco and N. Evans Generative models such as conditional variational auto-encoders (CVAEs) are capable of modelling complex data distributions via latent representation learning. This paper reports their application to ABE. CVAEs, form of directed, graphical models, are exploited to model the probability distribution of highband features conditioned on narrowband features.
1388 Using Automatic Speech Recognition and Speech Synthesis to Improve the Intelligibility of Cochlear Implant users in Reverberant Listening Environments K. Chu, L. Collins and B. Mainsah Thus, the current study investigated the performance of the previously proposed algorithm in multiple unseen environments.
1389 Speech Intelligibility Enhancement by Equalization for in-Car Applications E. Gentet, B. David, S. Denjean, G. Richard and V. Roussarie In this paper, we propose a speech intelligibility enhancement method for typical in-car applications in noisy environments.
1390 Maximum Likelihood Estimation of the Interference-Plus-Noise Cross Power Spectral Density Matrix for Own Voice Retrieval P. Hoang, Z. Tan, T. Lunner, J. M. de Haan and J. Jensen In this paper, a novel maximum likelihood (ML) estimator of the interference-plus-noise CPSD matrix is proposed.
1391 A Constrained Maximum Likelihood Estimator of Speech and Noise Spectra with Application to Multi-Microphone Noise Reduction A. Zahedi, M. S. Pedersen, J. ?stergaard, L. Bramsl?w, T. U. Christiansen and J. Jensen In this paper, we suggest a new estimation technique that tackles this issue by enforcing a power constraint on the estimation problem.
1392 CLCNET: Deep Learning-Based Noise Reduction for Hearing aids using Complex Linear Coding H. Schr?ter, T. Rosenkranz, A. N. Escalante-B, M. Aubreville and A. Maier To improve monaural speech enhancement in noisy environments, we propose CLCNet, a framework based on complex valued linear coding.
1393 A Time-Frequency Network with Channel Attention and Non-Local Modules for Artificial Bandwidth Extension Y. Dong et al. In this paper, we introduce a Time-Frequency Network (TFNet) with channel attention (CA) and non-local (NL) modules for ABE.
1394 Masking and Inpainting: A Two-Stage Speech Enhancement Approach for Low SNR and Non-Stationary Noise X. Hao et al. For better speech enhancement at the above scenarios, this paper proposes a two-stage approach that consists of binary masking and spectrogram inpainting.
1395 3-D Acoustic Modeling for Far-Field Multi-Channel Speech Recognition A. Purushothaman, A. Sreeram and S. Ganapathy In this paper, we propose to model the multi-channel signal directly using a convolutional neural network (CNN) based architecture which performs the joint acoustic modeling on the three dimensions of time, frequency and channel.
1396 Improving Reverberant Speech Training Using Diffuse Acoustic Simulation Z. Tang, L. Chen, B. Wu, D. Yu and D. Manocha We present an efficient and realistic geometric acoustic simulation approach for generating and augmenting training data in speech-related machine learning tasks.
1397 Low-Frequency Compensated Synthetic Impulse Responses For Improved Far-Field Speech Recognition Z. Tang, H. Meng and D. Manocha We propose a method for generating low-frequency compensated synthetic impulse responses that improve the performance of farfield speech recognition systems trained on artificially augmented datasets.
1398 Aipnet: Generative Adversarial Pre-Training of Accent-Invariant Networks for End-To-End Speech Recognition Y. Chen, Z. Yang, C. Yeh, M. Jain and M. L. Seltzer In this paper, our goal is to build a unified end-to-end speech recognition system that generalizes well across accents.
1399 Audio-Visual Recognition of Overlapped Speech for the LRS2 Dataset J. Yu et al. Motivated by the bimodal nature of human speech perception, this paper investigates the use of audio-visual technologies for overlapped speech recognition.
1400 Multi-Task Self-Supervised Learning for Robust Speech Recognition M. Ravanelli et al. This paper proposes PASE+, an improved version of PASE for robust speech recognition in noisy and reverberant environments.
1401 End-to-End Multi-Person Audio/Visual Automatic Speech Recognition O. Braga, T. Makino, O. Siohan and H. Liao We propose a fully differentiable A/V ASR model that is able to handle multiple face tracks in a video.
1402 End-to-End Automatic Speech Recognition Integrated with CTC-Based Voice Activity Detection T. Yoshimura, T. Hayashi, K. Takeda and S. Watanabe This paper integrates a voice activity detection (VAD) function with end-to-end automatic speech recognition toward an online speech interface and transcribing very long audio recordings.
1403 End-to-End Training of Time Domain Audio Separation and Recognition T. von Neumann et al. We here demonstrate how to combine a separation module based on a Convolutional Time domain Audio Separation Network (Conv-TasNet) with an E2E speech recognizer and how to train such a model jointly by distributing it over multiple GPUs or by approximating truncated back-propagation for the convolutional front-end.
1404 Improving Noise Robust Automatic Speech Recognition with Single-Channel Time-Domain Enhancement Network K. Kinoshita, T. Ochiai, M. Delcroix and T. Nakatani In this paper, we show that a single-channel time-domain denoising approach can significantly improve ASR performance, providing more than 30 % relative word error reduction over a strong ASR back-end on the real evaluation data of the single-channel track of the CHiME-4 dataset.
1405 A Practical Two-Stage Training Strategy for Multi-Stream End-to-End Speech Recognition R. Li, G. Sell, X. Wang, S. Watanabe and H. Hermansky In this work, we propose a practical two-stage training scheme.
1406 Multi-Scale Octave Convolutions for Robust Speech Recognition J. Rownicka, P. Bell and S. Renals We propose a multi-scale octave convolution layer to learn robust speech representations efficiently.
1407 Learning Noise Invariant Features Through Transfer Learning For Robust End-to-End Speech Recognition S. Zhang, C. Do, R. Doddipatla and S. Renals To address this, we propose transfer learning from a clean dataset (WSJ) to a noisy dataset (CHiME4) for connectionist temporal classification models.
1408 Improving Speech Recognition Using Consistent Predictions on Synthesized Speech G. Wang et al. In this work, we demonstrate that promoting consistent predictions in response to real and synthesized speech enables significantly improved speech recognition performance.
1409 Attention-Based ASR with Lightweight and Dynamic Convolutions Y. Fujita, A. S. Subramanian, M. Omachi and S. Watanabe In this paper, we propose to apply lightweight and dynamic convolution to E2E ASR as an alternative architecture to the self-attention to make the computational order linear.
1410 An Attention-Based Joint Acoustic and Text on-Device End-To-End Model T. N. Sainath, R. Pang, R. J. Weiss, Y. He, C. Chiu and T. Strohman In this work, we introduce a joint acoustic and text decoder (JATD) into the LAS decoder, which makes it possible to incorporate a much larger text corpus into training.
1411 Structured Sparse Attention for end-to-end Automatic Speech Recognition J. Xue, T. Zheng and J. Han In this paper, we present two sparse attention mechanisms for ASR tasks with long utterances, which try to improve the attention mechanism by introducing the sparse transformation.
1412 Rnn-Transducer with Stateless Prediction Network M. Ghodsi, X. Liu, J. Apfel, R. Cabrera and E. Weinstein We focus on the prediction network of the RNNT, since it is believed to be analogous to the Language Model (LM) in the classic ASR systems.
1413 Sequence-Level Consistency Training for Semi-Supervised End-to-End Automatic Speech Recognition R. Masumura, M. Ihori, A. Takashima, T. Moriya, A. Ando and Y. Shinohara This paper presents a novel semi-supervised end-to-end automatic speech recognition (ASR) method that employs consistency training with the use of unlabeled data.
1414 Independent Language Modeling Architecture for End-To-End ASR V. T. Pham et al. To address this problem, in this work, we propose a new architecture that separates the decoder subnet from the encoder output.
1415 Speaker-Aware Training of Attention-Based End-to-End Speech Recognition Using Neural Speaker Embeddings A. Rouhe, T. Kaseva and M. Kurimo We apply speaker-aware training to attention-based end-to-end speech recognition. We show that it can improve over a purely end-to-end baseline.
1416 Generating Synthetic Audio Data for Attention-Based Speech Recognition Systems N. Rossenbach, A. Zeyer, R. Schl?ter and H. Ney We extend state-of-the-art attention-based automatic speech recognition (ASR) systems with synthetic audio generated by a TTS system trained only on the ASR corpora itself.
1417 Correction of Automatic Speech Recognition with Transformer Sequence-To-Sequence Model O. Hrinchuk, M. Popova and B. Ginsburg In this work, we introduce a simple yet efficient post-processing model for automatic speech recognition.
1418 Exploring Pre-Training with Alignments for RNN Transducer Based End-to-End Speech Recognition H. Hu, R. Zhao, J. Li, L. Lu and Y. Gong In this work, we conversely leverage external alignments to seed the RNN-T model.
1419 Self-Training for End-to-End Speech Recognition J. Kahn, A. Lee and A. Hannun We revisit self-training in the context of end-to-end speech recognition.
1420 Toward Better Speaker Embeddings: Automated Collection of Speech Samples From Unknown Distinct Speakers M. Pham, Z. Li and J. Whitehill With the goal of training better embedding models, we devise an automatic pipeline for large-scale collection of speech samples from unique speakers that is significantly more automated than previous approaches.
1421 Channel Adversarial Training for Speaker Verification and Diarization C. Luu, P. Bell and S. Renals We propose a training strategy which aims to produce features that are invariant at the granularity of the recording or channel, a finer grained objective than dataset- or environment-invariance.
1422 Progressive Multi-Target Network Based Speech Enhancement with Snr-Preselection for Robust Speaker Diarization L. Sun, J. Du, X. Zhang, T. Gao, X. Fang and C. Lee In this paper, we design a novel front-end processing system for speaker diarization under realistic conditions with challenging background noises.
1423 Improved Large-Margin Softmax Loss for Speaker Diarisation Y. Fathullah, C. Zhang and P. C. Woodland Therefore, this paper introduces a general approach to the large-margin softmax loss without any approximations to improve the quality of speaker embeddings for diarisation.
1424 Speaker Diarization with Session-Level Speaker Embedding Refinement Using Graph Neural Networks J. Wang, X. Xiao, J. Wu, R. Ramamurthy, F. Rudzicz and M. Brudno In this work we present the first use of graph neural networks (GNNs) for the speaker diarization problem, utilizing a GNN to refine speaker embeddings locally using the structural information between speech segments inside each session.
1425 Overlap-Aware Diarization: Resegmentation Using Neural End-to-End Overlapped Speech Detection L. Bullock, H. Bredin and L. P. Garcia-Perera We address the problem of effectively handling overlapping speech in a diarization system.
1426 On the Importance of Vocal Tract Constriction for Speaker Characterization: The Whispered Speech Study R. K. Das and H. Li On the Importance of Vocal Tract Constriction for Speaker Characterization: The Whispered Speech Study
1427 Pyannote.Audio: Neural Building Blocks for Speaker Diarization H. Bredin et al. We introduce pyannote.audio, an open-source toolkit written in Python for speaker diarization.
1428 Speaker Embeddings Incorporating Acoustic Conditions for Diarization Y. Higuchi, M. Suzuki and G. Kurata We present our work on training speaker embeddings, especially effective for speaker diarization.
1429 Supervised Online Diarization with Sample Mean Loss for Multi-Domain Data E. Fini and A. Brutti In this paper we propose qualitative modifications to the model that significantly improve the learning efficiency and the overall diarization performance.
1430 Investigation of Specaugment for Deep Speaker Embedding Learning S. Wang, J. Rohdin, O. Plchot, L. Burget, K. Yu and J. Cernock? In this paper, we investigate the usage of SpecAugment for speaker verification tasks.
1431 Speaker-Invariant Affective Representation Learning via Adversarial Training H. Li, M. Tu, J. Huang, S. Narayanan and P. Georgiou In this paper, we propose a machine learning framework to obtain speech emotion representations by limiting the effect of speaker variability in the speech signals.
1432 Speech Sentiment Analysis via Pre-Trained Features from End-to-End ASR Models Z. Lu, L. Cao, Y. Zhang, C. Chiu and J. Fan In this paper, we propose to use pre-trained features from end-to-end ASR models to solve speech sentiment analysis as a down-stream task.
1433 Gender Differences on the Perception and Production of Utterances with Willingness and Reluctance in Chinese W. Li, J. W. Wong and J. Tu This study intends to explore the effects of gender differences on the perception and production of emotional intonation with willingness and reluctance.
1434 Hierarchical Attention Transfer Networks for Depression Assessment from Speech Z. Zhao, Z. Bao, Z. Zhang, N. Cummins, H. Wang and B. Schuller In this paper, we propose a novel crosstask approach which transfers attention mechanisms from speech recognition to aid depression severity measurement.
1435 Detecting Emotion Primitives from Speech and Their Use in Discerning Categorical Emotions V. Kowtha et al. This work investigated a long-shot-term memory (LSTM) network and a time convolution – LSTM (TC-LSTM) to detect primitive emotion attributes such as valence, arousal, and dominance, from speech.
1436 X-Vectors Meet Emotions: A Study On Dependencies Between Emotion and Speaker Recognition R. Pappagari, T. Wang, J. Villalba, N. Chen and N. Dehak In this work, we explore the dependencies between speaker recognition and emotion recognition.
1437 Speech Emotion Recognition with Local-Global Aware Deep Representation Learning J. Liu, Z. Liu, L. Wang, L. Guo and J. Dang In this paper, we propose a local-global aware deep representation learning system that mainly includes two modules.
1438 Multi-Head Attention for Speech Emotion Recognition with Auxiliary Learning of Gender Recognition A. Nediyanchath, P. Paramasivam and P. Yenigalla The paper presents a Multi-Head Attention deep learning network for Speech Emotion Recognition (SER) using Log mel-Filter Bank Energies (LFBE) spectral features as the input.
1439 Generating and Protecting Against Adversarial Attacks for Deep Speech-Based Emotion Recognition Models Z. Ren, A. Baird, J. Han, Z. Zhang and B. Schuller With this in mind, in this study, we aim to train deep models to defending against non-targeted white-box adversarial attacks.
1440 Deep Encoded Linguistic and Acoustic Cues for Attention Based End to End Speech Emotion Recognition S. Bhosale, R. Chakraborty and S. K. Kopparapu An End-to-End model with convolutional layers and multi-head self attention mechanism is proposed for Speech Emotion Recognition (SER) task.
1441 Multi-Conditioning and Data Augmentation Using Generative Noise Model for Speech Emotion Recognition in Noisy Conditions U. Tiwari, M. Soni, R. Chakraborty, A. Panda and S. K. Kopparapu In this paper, to address the robustness aspect of the SER in additive noise scenarios, we propose multi-conditioning and data augmentation using an utterance level parametric Generative noise model.
1442 A Self-Attentive Emotion Recognition Network H. Partaourides, K. Papadamou, N. Kourtellis, I. Leontiades and S. Chatzis In this work, we introduce a novel attention mechanism capable of inferring the immensity of the effect of each past utterance on the current speaker emotional state.
1443 Efficient Shallow Wavenet Vocoder Using Multiple Samples Output Based on Laplacian Distribution and Linear Prediction P. L. Tobing, Y. Wu, T. Hayashi, K. Kobayashi and T. Toda In this work, we tackle this issue by proposing a simple implementation scheme of segment output modeling, that can be easily extended into other neural vocoders, where the Laplacian distribution parameters of multiple samples are estimated simultaneously.
1444 Flow-TTS: A Non-Autoregressive Network for Text to Speech Based on Flow C. Miao, S. Liang, M. Chen, J. Ma, S. Wang and J. Xiao In this work, we propose Flow-TTS, a non-autoregressive end-to-end neural TTS model based on generative flow.
1445 WaveFFJORD: FFJORD-Based Vocoder for Statistical Parametric Speech Synthesis N. Wu and Z. Ling Inspired by WaveGlow, in this paper, we propose WaveFFJORD, a neural vocoder that can synthesize speech waveforms from acoustic features, by combining FFJORD and WaveNet.
1446 Improving LPCNET-Based Text-to-Speech with Linear Prediction-Structured Mixture Density Network M. Hwang, E. Song, R. Yamamoto, F. Soong and H. Kang In this paper, we propose an improved LPCNet vocoder using a linear prediction (LP)-structured mixture density network (MDN).
1447 Disentangling Timbre and Singing Style with Multi-Singer Singing Synthesis System J. Lee, H. Choi, J. Koo and K. Lee In this study, we define the identity of the singer with two independent concepts – timbre and singing style – and propose a multi-singer singing synthesis system that can model them separately.
1448 Sequence-to-Sequence Singing Synthesis Using the Feed-Forward Transformer M. Blaauw and J. Bonada We propose a sequence-to-sequence singing synthesizer, which avoids the need for training data with pre-aligned phonetic and acoustic features.
1449 Korean Singing Voice Synthesis Based on Auto-Regressive Boundary Equilibrium Gan S. Choi, W. Kim, S. Park, S. Yong and J. Nam In this paper, we propose a Korean singing voice synthesis system that addresses the issues using an auto-regressive algorithm that generates spectrogram with the boundary equilibrium GAN objective.
1450 Fast and High-Quality Singing Voice Synthesis System Based on Convolutional Neural Networks K. Nakamura, S. Takaki, K. Hashimoto, K. Oura, Y. Nankaku and K. Tokuda The present paper describes singing voice synthesis based on convolutional neural networks (CNNs).
1451 Hybrid Neural-Parametric F0 Model for Singing Synthesis J. Bonada and M. Blaauw We propose a novel hybrid neural-parametric fundamental frequency generation model for singing voice synthesis.
1452 Utterance-Level Sequential Modeling for Deep Gaussian Process Based Speech Synthesis Using Simple Recurrent Unit T. Koriyama and H. Saruwatari This paper presents a deep Gaussian process (DGP) model with a recurrent architecture for speech sequence modeling.
1453 Emotional Speech Synthesis with Rich and Granularized Control S. Um, S. Oh, K. Byun, I. Jang, C. Ahn and H. Kang This paper proposes an effective emotion control method for an end-to-end text-to-speech (TTS) system.
1454 Towards Unsupervised Speech Recognition and Synthesis with Quantized Speech Representation Learning A. H. Liu, T. Tu, H. Lee and L. Lee In this paper we propose a Sequential Representation Quantization AutoEncoder (SeqRQ-AE) to learn from primarily unpaired audio data and produce sequences of representations very close to phoneme sequences of speech utterances.
1455 An Empirical Study of Conv-Tasnet B. Kadioglu, M. Horgan, X. Liu, J. Pons, D. Darcy and V. Kumar In this paper, we conduct an empirical study of Conv-TasNet and propose an enhancement to the encoder/decoder that is based on a (deep) non-linear variant of it.
1456 Mask-Dependent Phase Estimation for Monaural Speaker Separation Z. Ni and M. I. Mandel In this paper, we propose a simple yet effective phase estimation network that predicts the phase of the clean speech based on a T-F mask predicted by a chimera++ network.
1457 Joint Phoneme Alignment and Text-Informed Speech Separation on Highly Corrupted Speech K. Schulze-Forster, C. S. J. Doire, G. Richard and R. Badeau We propose to perform text-informed speech-music separation and phoneme alignment jointly using recurrent neural networks and the attention mechanism.
1458 Single-Channel Speech Separation Integrating Pitch Information Based on a Multi Task Learning Framework X. Li, R. Liu, T. Song, X. Wu and J. Chen In this paper, we aimed to combine speech separation and pitch tracking together to let them benefit from each other.
1459 Continuous Speech Separation: Dataset and Analysis Z. Chen et al. In this paper, we define continuous speech separation (CSS) as a task of generating a set of non-overlapped speech signals from a continuous audio stream that contains multiple utterances that are partially overlapped by a varying degree.
1460 The Sound of My Voice: Speaker Representation Loss for Target Voice Separation S. Mun, S. Choe, J. Huh and J. S. Chung In this paper, we propose a new loss function using speaker content representation for audio source separation, and we call it speaker representation loss.
1461 Speaker-Aware Target Speaker Enhancement by Jointly Learning with Speaker Embedding Extraction X. Ji et al. In this paper, we propose a novel training framework which jointly learns the speaker-conditioned target speaker extraction model and its associated speaker embedding model.
1462 Far-Field Location Guided Target Speech Extraction Using End-to-End Speech Recognition Objectives A. S. Subramanian et al. This paper proposes a method to jointly optimize a location guided target speech extraction module along with a speech recognition module only with ASR error minimization criteria.
1463 A Study of Child Speech Extraction Using Joint Speech Enhancement and Separation in Realistic Conditions X. Wang, J. Du, A. Cristia, L. Sun and C. Lee In this paper, we design a novel joint framework of speech enhancement and speech separation for child speech extraction in realistic conditions, targeting the problem of extracting child speech from daily conversations in BabyTrain mega corpus.
1464 An Analysis of Speech Enhancement and Recognition Losses in Limited Resources Multi-Talker Single Channel Audio-Visual ASR L. Pasa, G. Morrone and L. Badino In this paper, we analyzed how audio-visual speech enhancement can help to perform the ASR task in a cocktail party scenario.
1465 Deep Audio-Visual Speech Separation with Attention Mechanism C. Li and Y. Qian In this paper, we explore a better strategy to utilize visual representations with the attention mechanism.
1466 Enhancing End-to-End Multi-Channel Speech Separation Via Spatial Feature Learning R. Gu et al. In this work, we propose an integrated architecture for learning spatial features directly from the multi-channel speech waveforms within an end-to-end speech separation framework.
1467 Detection and Analysis of T/D Deletion in Librispeech J. Yuan, H. Lin and Y. Liu In this study we developed a new method for automatic identification of t/d deletion.
1468 Prediction of Voicing and the F0 Contour from Electromagnetic Articulography Data for Articulation-to-Speech Synthesis S. Stone, P. Schmidt and P. Birkholz This paper examines to what extent the f0 contour of an utterance can be predicted from such supraglottal articulation data.
1469 A Comparative Study of Estimating Articulatory Movements from Phoneme Sequences and Acoustic Features A. Singh, A. Illa and P. K. Ghosh In this work, we estimate articulatory movements from three different input representations: R1) acoustic signal, R2) phoneme sequence, R3) phoneme sequence with timing information.
1470 Automatic Vocal Tractlandmark Tracking in Rtmri Using Fully Convolutional Networks and Kalman Filter S. Asadiabadi and E. Erzin In this work, we present an algorithm for robust detection of keypoints on the vocal tract in rtMRI sequences using fully convolutional networks (FCN) via a heatmap regression approach.
1471 Speech-Based Parameter Estimation of an Asymmetric Vocal Fold Oscillation Model and its Application in Discriminating Vocal Fold Pathologies W. Zhao and R. Singh In this paper, we present a novel and alternate way to determine vocal fold model parameters from the speech signal.
1472 End-to-End Articulatory Modeling for Dysarthric Articulatory Attribute Detection Y. Lin, L. Wang, J. Dang, S. Li and C. Ding In this study, we focus on detecting articulatory attribute errors for dysarthric patients with cerebral palsy (CP) or amyotrophic lateral sclerosis (ALS).
1473 Vocal Tract Articulatory Contour Detection in Real-Time Magnetic Resonance Images Using Spatio-Temporal Context S. A. Hebbar, R. Sharma, K. Somandepalli, A. Toutios and S. Narayanan In this work, we propose to use information from temporal dynamics together with the spatial structure to detect the articulatory boundaries in rtMRI videos.
1474 Retrieving Vocal-Tract Resonance and anti-Resonance From High-Pitched Vowels Using a Rahmonic Subtraction Technique Z. Zhang, K. Honda and J. Wei In this study, a new cepstral method is developed combined with a refined rahmonic subtraction technique (RS-CEPS) to extract spectral envelopes excited by glottal noise sources.
1475 Epoch Estimation from a Speech Signal Using Gammatone Wavelets in a Scattering Network P. Kulkarni, J. Sadasivan, A. Adiga and C. S. Seelamantula In this paper, we propose a new technique that employs a Gammatone wavelet filterbank and compute a scattering sequence whose local maxima define the candidate epochs in the speech signal.
1476 Study of Closed Phase Resonance Bandwidths for Oral and Nasal Tracts Using Zero Time Windowing H. D. Abbas, R. Prasad, B. T. Nellore and S. V. Gangashetty Study of Closed Phase Resonance Bandwidths for Oral and Nasal Tracts Using Zero Time Windowing
1477 Algorithmic Exploration of American English Dialects A. Aks?nova, A. Bruguier, A. Ritchart-Scott and U. Mendlovic In this paper, we use a novel algorithmic approach to explore dialectal variation in American English speech.
1478 Comparison of Glottal Closure Instants Detection Algorithms for Emotional Speech S. R. Kadiri, P. Alku and B. Yegnanarayana In this paper, we compare six GCI detection algorithms using emotional speech and known evaluation metrics.
1479 Unsupervised Speaker Adaptation Using Attention-Based Speaker Memory for End-to-End ASR L. Sari, N. Moritz, T. Hori and J. L. Roux We propose an unsupervised speaker adaptation method inspired by the neural Turing machine for end-to-end (E2E) automatic speech recognition (ASR).
1480 L-Vector: Neural Label Embedding for Domain Adaptation Z. Meng et al. We propose a novel neural label embedding (NLE) scheme for the domain adaptation of a deep neural network (DNN) acoustic model with unpaired data samples from source and target domains.
1481 Acoustic Model Adaptation for Presentation Transcription and Intelligent Meeting Assistant Systems Y. Huang and Y. Gong We present our solution for unsupervised rapid speaker adaptation in a state-of-art presentation and intelligent meeting transcription system.
1482 Using Personalized Speech Synthesis and Neural Language Generator for Rapid Speaker Adaptation Y. Huang, L. He, W. Wei, W. Gale, J. Li and Y. Gong We propose to use the personalized speech synthesis and the neural language generator to synthesize content relevant personalized speech for rapid speaker adaptation.
1483 Attention-Based Gated Scaling Adaptive Acoustic Model for CTC-Based Speech Recognition F. Ding, W. Guo, L. Dai and J. Du In this paper, we propose a novel adaptive technique that uses an attention-based gated scaling (AGS) scheme to improve deep feature learning for connectionist temporal classification (CTC) acoustic modeling.
1484 Adaptive Knowledge Distillation Based on Entropy K. Kwon, H. Na, H. Lee and N. S. Kim In this paper, we propose an entropy based KD training, which utilizes the teacher model labels with lower entropy at a larger rate among the various teacher models.
1485 Unsupervised Pretraining Transfers Well Across Languages M. Rivi?re, A. Joulin, P. Mazar? and E. Dupoux In this work, we investigate whether unsupervised pretraining transfers well across languages.
1486 Incremental Semi-Supervised Learning for Multi-Genre Speech Recognition B. Khonglah, S. Madikeri, S. Dey, H. Bourlard, P. Motlicek and J. Billa In this work, we explore a data scheduling strategy for semi-supervised learning (SSL) for acoustic modeling in automatic speech recognition.
1487 Source Domain Data Selection for Improved Transfer Learning Targeting Dysarthric Speech Recognition F. Xiong, J. Barker, Z. Yue and H. Christensen This paper presents an improved transfer learning framework applied to robust personalised speech recognition models for speakers with dysarthria.
1488 Study of Formant Modification for Children ASR H. Kumar Kathania, S. Reddy Kadiri, P. Alku and M. Kurimo In this paper, we propose a formant modification method to mitigate differences between adults? and children?s speech and to improve the performance of ASR for children.
1489 Pseudo Likelihood Correction Technique for Low Resource Accented ASR A. Rajpal, A. R. MV, C. Yarra, R. Aggarwal and P. K. Ghosh In order to improve the performance of a native English ASR on non-native English data, we, in this work, propose a DNN-based pseudo-likelihood correction (PLC) technique, in which a non-native pseudo-likelihood vector is mapped to match its native counterpart.
1490 Libri-Adapt: a New Speech Dataset for Unsupervised Domain Adaptation A. Mathur, F. Kawsar, N. Berthouze and N. D. Lane This paper introduces a new dataset, Libri-Adapt, to support unsupervised domain adaptation research on speech recognition models.
1491 Mining Effective Negative Training Samples for Keyword Spotting J. Hou, Y. Shi, M. Ostendorf, M. Hwang and L. Xie To address the problem, we propose an innovative algorithm, Regional Hard-Example (RHE) mining, to find effective negative training samples, in order to control the ratio of negative vs. positive data.
1492 Multi-Task Learning for Voice Trigger Detection S. Sigtia, P. Clark, R. Haynes, H. Richards and J. Bridle In this study, we address two major challenges. The first is that the detectors are deployed in complex acoustic environments with external noise and loud playback by the device itself. Secondly, collecting training examples for a specific keyword or trigger phrase is challenging resulting in a scarcity of trigger phrase specific training data.
1493 Small-Footprint Keyword Spotting on Raw Audio Data with Sinc-Convolutions S. Mittermaier, L. K?rzinger, B. Waschneck and G. Rigoll Compared to previous publications, we took additional steps to reduce power and memory consumption without reducing classification accuracy.
1494 Lattice-Based Improvements for Voice Triggering Using Graph Neural Networks P. Dighe et al. In this paper, we address the task of false trigger mitigation (FTM) using a novel approach based on analyzing automatic speech recognition (ASR) lattices using graph neural networks (GNN).
1495 Integration of Multi-Look Beamformers for Multi-Channel Keyword Spotting X. Ji, M. Yu, J. Chen, J. Zheng, D. Su and D. Yu In this paper, we propose integrating multiple beamformed signals and a microphone signal as input to an end-to-end KWS model and leveraging the attention mechanism to dynamically tune the model?s attention to the reliable input sources.
1496 Fast Lattice-Free Keyword Filtering for Accelerated Spoken Term Detection J. Wintrode and J. Wilkes We present a novel set of keyword detection techniques to accelerate spoken term detection for known queries with minimal loss in accuracy.
1497 Training Keyword Spotters with Limited and Synthesized Speech Data J. Lin, K. Kilgour, D. Roblek and M. Sharifi In this paper, we explore the effectiveness of synthesized speech data in training small, spoken term detection models of around 400k parameters.
1498 Towards Data-Efficient Modeling for Wake Word Spotting Y. Gao, Y. Mishchenko, A. Shah, S. Matsoukas and S. Vitaladevuni In this paper we present data-efficient solutions to address the challenges in WW modeling, such as domain-mismatch, noisy conditions, limited annotation, etc.
1499 Adaptation of RNN Transducer with Text-To-Speech Technology for Keyword Spotting E. Sharma et al. With a goal to improve the KWS performance for these keywords without having to collect additional natural speech data, we explore Text-To-Speech (TTS) technology to synthetically generate training data for such keywords.
1500 Crnn-Ctc Based Mandarin Keywords Spotting H. Yan, Q. He and W. Xie In this paper, we propose an end-to-end Mandarin KWS system using Convolutional Recurrent Neural Network with the Connectionist Temporal Classification (CTC) loss function (CRNN-CTC).
1501 Unsupervised Neural Mask Estimator for Generalized Eigen-Value Beamforming Based Asr R. Kumar, A. Sreeram, A. Purushothaman and S. Ganapathy In this paper, we attempt to move away from the requirements of having supervised clean recordings for training the mask estimator.
1502 Spatial Attention for Far-Field Speech Recognition with Deep Beamforming Neural Networks W. He, L. Lu, B. Zhang, J. Mahadeokar, K. Kalgaonkar and C. Fuegen In this paper, we introduce spatial attention for refining the information in multi-direction neural beamformer for far-field automatic speech recognition.
1503 Tensor-To-Vector Regression for Multi-Channel Speech Enhancement Based on Tensor-Train Network J. Qi, H. Hu, Y. Wang, C. H. Yang, S. Marco Siniscalchi and C. Lee We propose a tensor-to-vector regression approach to multi-channel speech enhancement in order to address the issue of input size explosion and hidden-layer size expansion.
1504 Truth-to-Estimate Ratio Mask: A Post-Processing Method for Speech Enhancement Direct at Low Signal-to-Noise Ratios B. Chen, H. Wang, Y. Wei and R. H. Y. So This study proposes a bi-directional recurrent neural network (Bi-RNN) post-processing method for speech enhancement (SE) at low signal-to noise ratios (SNR).
1505 Geometry Constrained Progressive Learning for Lstm-Based Speech Enhancement X. Tang, J. Du, L. Chai, Y. Wang, Q. Wang and C. Lee In this paper, we incorporate two kinds of geometric constraints among these targets into the objective function to help LSTM achieve better training.
1506 Using Separate Losses for Speech and Noise in Mask-Based Speech Enhancement Z. Xu, S. Elshamy and T. Fingscheidt In this paper, we propose a novel components loss (CL) for the training of neural networks for speech enhancement.
1507 Stable Training of Dnn for Speech Enhancement Based on Perceptually-Motivated Black-Box Cost Function M. Kawanaka, Y. Koizumi, R. Miyazaki and K. Yatabe To overcome this problem, we propose to use stabilization techniques borrowed from reinforcement learning.
1508 A Robust Audio-Visual Speech Enhancement Model W. Wang, C. Xing, D. Wang, X. Chen and F. Sun In this paper, we present a safe AVSE approach that can make the visual stream contribute to audio speech enhancment(ASE) safely in conditions of various SNRs by late fusion.
1509 Robust Unsupervised Audio-Visual Speech Enhancement Using a Mixture of Variational Autoencoders M. Sadeghi and X. Alameda-Pineda In this paper, we propose a robust unsupervised audio-visual speech enhancement method based on a per-frame VAE mixture model.
1510 AV(SE)2: Audio-Visual Squeeze-Excite Speech Enhancement M. L. Iuzzolino and K. Koishida We propose a new mechanism for audio-visual (AV) fusion that leverages a cross-modal squeeze-excitation (SE) block for speech enhancement: AV(SE)2.
1511 Spectrograms Fusion with Minimum Difference Masks Estimation for Monaural Speech Dereverberation H. Shi, L. Wang, M. Ge, S. Li and J. Dang In this study, we design the minimum difference masks (MDMs) to classify the time-frequency (T-F) bins in spectrograms according to the nearest distances from labels.
1512 A Return to Dereverberation in the Frequency Domain Using a Joint Learning Approach Y. Li and D. S. Williamson In this paper, we investigate whether deverberation can be effectively performed in the frequency-domain by estimating the complex frequency response of a room impulse response.
1513 In-Domain and Out-of-Domain Data Augmentation to Improve Children?s Speaker Verification System in Limited Data Scenario S. Shahnawazuddin, W. Ahmad, N. Adiga and A. Kumar In this paper, we present our efforts towards developing a robust automatic speaker verification (ASV) system for children when the domain-specific data is limited.
1514 Jhu-HLTCOE System for the Voxsrc Speaker Recognition Challenge D. Garcia-Romero, A. McCree, D. Snyder and G. Sell This paper describes our submission to this challenge where we have explored x-vector extractor topologies, classification head alternatives, data augmentation, and angular margin penalty.
1515 Detection of Speech Events and Speaker Characteristics through Photo-Plethysmographic Signal Neural Processing G. C?mbara, J. Luque and M. Farr?s In this work, we explore several end-to-end convolutional neural network architectures for detection of human?s characteristics such as gender or person identity.
1516 XMU-TS Systems for NIST SRE19 CTS Challenge H. Lu, J. Zhou, M. Zhao, W. Lei, Q. Hong and L. Li In this paper, we present our submitted XMU-TS system for NIST SRE19 CTS Challenge.
1517 I-Vector Transformation Using K-Nearest Neighbors for Speaker Verification U. Khan, M. India and J. Hernando In this work, we propose a post processing of i-vectors using a Deep Neural Network (DNN) to transform i-vectors into a new speaker vector representation.
1518 H-Vectors: Utterance-Level Speaker Embedding Using a Hierarchical Attention Model Y. Shi, Q. Huang and T. Hain In this paper, a hierarchical attention network is proposed to generate utterance-level embeddings (H-vectors) for speaker identification and verification.
1519 Feature Enhancement with Deep Feature Losses for Speaker Verification S. Kataria, P. S. Nidadavolu, J. Villalba, N. Chen, P. Garc?a-Perera and N. Dehak We propose to use Deep Feature Loss which optimizes the enhancement network in the hidden activation space of a pre-trained auxiliary speaker embedding network.
1520 Combining Deep Embeddings of Acoustic and Articulatory Features for Speaker Identification Q. Hong, C. Wu, H. Wang and C. Huang In this study, deep embedding of acoustic and articulatory features are combined for speaker identification.
1521 Bayesian Estimation of Plda with Noisy Training Labels, with Applications to Speaker Verification B. J. Borgstr?m and P. Torres-Carrasquillo This paper proposes a method for Bayesian estimation of probabilistic linear discriminant analysis (PLDA) when training labels are noisy.
1522 Unsupervised Feature Enhancement for Speaker Verification P. S. Nidadavolu, S. Kataria, J. Villalba, P. Garc?a-Perera and N. Dehak We developed an unsupervised feature enhancement approach in log-filter bank space with the end goal of improving speaker verification performance.
1523 CN-Celeb: A Challenging Chinese Speaker Recognition Dataset Y. Fan et al. In this paper, we present CN-Celeb, a large-scale speaker recognition dataset collected �in the wild�.
1524 HI-MIA: A Far-Field Text-Dependent Speaker Verification Database and the Baselines X. Qin, H. Bu and M. Li This paper presents a far-field text-dependent speaker verification database named HI-MIA.
1525 End-to-End Code-Switching TTS with Cross-Lingual Language Model X. Zhou, X. Tian, G. Lee, R. K. Das and H. Li In this paper, we propose to incorporate crosslingual word embedding into an end-to-end TTS system, to improve the voice rendering.
1526 Code-Switched Speech Synthesis Using Bilingual Phonetic Posteriorgram with Only Monolingual Corpora Y. Cao et al. In this paper, we present a bilingual phonetic posteriorgram (PPG) based CS speech synthesizer using only monolingual corpora.
1527 Generating Multilingual Voices Using Speaker Space Translation Based on Bilingual Speaker Data S. Maiti, E. Marchi and A. Conkie We present progress towards bilingual Text-to-Speech which is able to transform a monolingual voice to speak a second language while preserving speaker voice quality.
1528 Speaker Adaptation of a Multilingual Acoustic Model for Cross-Language Synthesis I. Himawan, S. Aryal, I. Ouyang, S. Kang, P. Lanchantin and S. King In the current work, our objective is to synthesize speech in different languages using the target speaker?s voice, regardless of the language of their data.
1529 Semi-Supervised Speaker Adaptation for End-to-End Speech Synthesis with Pretrained Models K. Inoue, S. Hara, M. Abe, T. Hayashi, R. Yamamoto and S. Watanabe To make use of such accessible data, the proposed method leverages the recent great success of state-of-the-art end-to-end automatic speech recognition (ASR) systems and obtains corresponding transcriptions from pretrained ASR models.
1530 BOFFIN TTS: Few-Shot Speaker Adaptation by Bayesian Optimization H. B. Moss, V. Aggarwal, N. Prateek, J. Gonz?lez and R. Barra-Chicote We present BOFFIN TTS (Bayesian Optimization For FIne-tuning Neural Text To Speech), a novel approach for few-shot speaker adaptation.
1531 Semi-Supervised Learning Based on Hierarchical Generative Models for End-to-End Speech Synthesis T. Fujimoto, S. Takaki, K. Hashimoto, K. Oura, Y. Nankaku and K. Tokuda This paper proposes a general framework of semi-supervised learning based on hierarchical generative models and adapts it to a Japanese end-to-end text-to-speech (TTS) system.
1532 Breathing and Speech Planning in Spontaneous Speech Synthesis ?. Sz?kely, G. E. Henter, J. Beskow and J. Gustafson To address this, we first propose training stochastic TTS on a corpus of overlapping breath-group bigrams, to take context into account. Next, we introduce an unsupervised automatic annotation of likely-disfluent breath events, through a product-of-experts model that combines the output of two breath- event predictors, each using complementary information and operating in opposite directions.
1533 Espnet-TTS: Unified, Reproducible, and Integratable Open Source End-to-End Text-to-Speech Toolkit T. Hayashi et al. This paper introduces a new end-to-end text-to-speech (E2E-TTS) toolkit named ESPnet-TTS, which is an extension of the open-source speech processing toolkit ESPnet.
1534 Extracting Unit Embeddings Using Sequence-To-Sequence Acoustic Models for Unit Selection Speech Synthesis X. Zhou, Z. Ling and L. Dai This paper presents a method of using the intermediate representations between linguistic and acoustic features in a Tacotron model to derive the cost functions for unit selection speech synthesis.
1535 Audio-Assisted Image Inpainting for Talking Faces A. Koumparoulis, G. Potamianos, S. Thomas and E. da Silva Morais The goal of our work is to complete missing areas of images of talking faces, exploiting information from both the visual and audio modalities.
1536 Libri-Light: A Benchmark for ASR with Limited or No Supervision J. Kahn et al. We introduce a new collection of spoken English audio suitable for training speech recognition systems under limited or no supervision.
1537 A Comprehensive Study of Residual CNNS for Acoustic Modeling in ASR V. Bozheniuk, A. Zeyer, R. Schl?ter and H. Ney Apart from recognition performance, we focus on the training and evaluation speed and provide a time-efficient setup for CNNs.
1538 Layer-Normalized LSTM for Hybrid-Hmm and End-To-End ASR M. Zeineldeen, A. Zeyer, R. Schl?ter and H. Ney We explore various LN long short-term memory (LSTM) recurrent neural networks (RNN) variants by applying LN to different parts of the internal recurrency of LSTMs.
1539 Small Energy Masking for Improved Neural Network Training for End-To-End Speech Recognition C. Kim, K. Kim and S. Reddy Indurthi In this paper, we present a Small Energy Masking (SEM) algorithm, which masks inputs having values below a certain threshold.
1540 Improving Sequence-To-Sequence Speech Recognition Training with On-The-Fly Data Augmentation T. Nguyen, S. St?ker, J. Niehues and A. Waibel In this paper we examine the influence of three data augmentation methods on the performance of two S2S model architectures.
1541 Effectiveness of Self-Supervised Pre-Training for ASR A. Baevski and A. Mohamed We compare self-supervised representation learning algorithms which either explicitly quantize the audio data or learn representations without quantization.
1542 High-Accuracy and Low-Latency Speech Recognition with Two-Head Contextual Layer Trajectory LSTM Model J. Li et al. In this paper, we detail our recent efforts to improve conventional hybrid LSTM acoustic models for high-accuracy and low-latency automatic speech recognition.
1543 Dfsmn-San with Persistent Memory Model for Automatic Speech Recognition Z. You, D. Su, J. Chen, C. Weng and D. Yu In this paper, we try to investigate whether even more information beyond the whole utterance level can be exploited and beneficial.
1544 Dynamic Temporal Residual Learning for Speech Recognition J. Xie, R. Yan, S. Xiao, L. Peng, M. T. Johnson and W. Zhang This paper proposes a novel dynamic temporal residual learning mechanism for LSTM networks to better explore temporal dependencies in sequential data.
1545 E2E-SINCNET: Toward Fully End-To-End Speech Recognition T. Parcollet, M. Morchid and G. Linar?s In this paper, we propose the E2E-SincNet, a novel fully E2E ASR model that goes from the raw waveform to the text transcripts by merging two recent and powerful paradigms: SincNet and the joint CTC-attention training scheme.
1546 Speaker Augmentation for Low Resource Speech Recognition C. Du and K. Yu In this paper, we propose a novel speaker augmentation approach which can synthesize data with sufficient speaker and text diversity.
1547 CGCNN: Complex Gabor Convolutional Neural Network on Raw Speech P. No?, T. Parcollet and M. Morchid We propose to combine the complex Gabor filter with complex-valued deep neural networks to replace usual CNN weights kernels, to fully take advantage of its optimal time-frequency resolution and of the complex domain.
1548 One-Shot Voice Conversion Using Star-Gan R. Wang, Y. Ding, L. Li and C. Fan Our efforts are made on one-shot voice conversion where the target speaker is unseen in training dataset or both source and target speakers are unseen in the training dataset.
1549 One-Shot Voice Conversion by Vector Quantization D. Wu and H. Lee In this paper, we propose a vector quantization (VQ) based one-shot voice conversion (VC) approach without any supervision on speaker label.
1550 Neutral to Lombard Speech Conversion with Deep Learning E. Gentet, B. David, S. Denjean, G. Richard and V. Roussarie In this paper, we propose several approaches for neutral to Lombard speech conversion.
1551 End-To-End Voice Conversion Via Cross-Modal Knowledge Distillation for Dysarthric Speech Reconstruction D. Wang et al. Inspired by the success of sequence-to-sequence (seq2seq) based text-to-speech (TTS) synthesis and knowledge distillation (KD) techniques, this paper proposes a novel end-to-end voice conversion (VC) method to tackle the reconstruction task.
1552 Pitchnet: Unsupervised Singing Voice Conversion with Pitch Adversarial Network C. Deng, C. Yu, H. Lu, C. Weng and D. Yu In this paper, we propose to advance the existing unsupervised singing voice conversion method proposed in [1] to achieve more accurate pitch translation and flexible pitch manipulation.
1553 An Improved Frame-Unit-Selection Based Voice Conversion System Without Parallel Training Data F. Xie et al. An Improved Frame-Unit-Selection Based Voice Conversion System Without Parallel Training Data
1554 Voice Conversion with Transformer Network R. Liu, X. Chen and X. Wen This paper describes an end-to-end voice conversion system, which involves three main ideas: transformer, context preservation mechanisms, and model adaptation.
1555 Mspec-Net : Multi-Domain Speech Conversion Network H. Malaviya, J. Shah, M. Patel, J. Munshi and H. A. Patil In this paper, we present a multi-domain speech conversion technique by proposing a Multi-domain Speech Conversion Network (MSpeC-Net) architecture for solving the less-explored area of Non-Audible Murmur-to-SPeeCH (NAM2-SPCH) conversion.
1556 Multi-Speaker and Multi-Domain Emotional Voice Conversion Using Factorized Hierarchical Variational Autoencoder M. Elgaar, J. Park and S. W. Lee This study exploits the FHVAE pipeline to produce disentangled representations of emotion, making it possible to greatly facilitate emotional voice conversion.We propose three versions of algorithms for improving the quality of disentangled representation and audio synthesis.
1557 Emotional Voice Conversion Using Multitask Learning with Text-To-Speech T. Kim, S. Cho, S. Choi, S. Park and S. Lee In this study, a voice converter that utilizes multitask learning with text-to-speech (TTS) is presented.
1558 Effective Wavenet Adaptation for Voice Conversion with Limited Data H. Du, X. Tian, L. Xie and H. Li In this paper, we propose a WaveNet adaptation method that effectively reduces the need of adaptation data.
1559 Lifter Training and Sub-Band Modeling for Computationally Efficient and High-Quality Voice Conversion Using Spectral Differentials T. Saeki, Y. Saito, S. Takamichi and H. Saruwatari In this paper, we propose computationally efficient and high-quality methods for statistical voice conversion (VC) with direct waveform modification based on spectral differentials.
1560 Improving Proper Noun Recognition in End-To-End Asr by Customization of the Mwer Loss Criterion C. Peyser, T. N. Sainath and G. Pundak In this paper, we instead build on recent advances in minimum word error rate (MWER) training to develop two new loss criteria that specifically emphasize proper noun recognition.
1561 Neural Lattice Search for Speech Recognition R. Ma, H. Li, Q. Liu, L. Chen and K. Yu In this paper, we address these problems with an end-to-end model for accurately extracting the best hypothesis from the word lattice.
1562 Deliberation Model Based Two-Pass End-To-End Speech Recognition K. Hu, T. N. Sainath, R. Pang and R. Prabhavalkar In this work, we propose to attend to both acoustics and first-pass hypotheses using a deliberation network.
1563 Alignment-Length Synchronous Decoding for RNN Transducer G. Saon, Z. T?ske and K. Audhkhasi We present a beam decoding strategy for recurrent neural network transducers which has the characteristic that all competing hypotheses within the beam have the same alignment length (number of output symbols plus BLANK symbols).
1564 Incorporating Written Domain Numeric Grammars into End-To-End Contextual Speech Recognition Systems for Improved Recognition of Numeric Sequences B. Haynor and P. S. Aleksic We propose a modular and scalable solution for improved recognition of numeric sequences.
1565 LSTM-Based One-Pass Decoder for Low-Latency Streaming J. Jorge et al. In this paper we present a novel streaming decoder that includes a bidirectional LSTM acoustic model as well as an unidirectional LSTM language model to perform the decoding efficiently while keeping the performance comparable to that of an off-line setup.
1566 Multistate Encoding with End-To-End Speech RNN Transducer Network Z. Wu, B. Li, Y. Zhang, P. S. Aleksic and T. N. Sainath In this paper, we propose a technique for incorporating contextual signals, such as intelligent assistant device state or dialog state, directly into RNN-T models.
1567 Neural Oracle Search on N-BEST Hypotheses E. Variani, T. Chen, J. Apfel, B. Ramabhadran, S. Lee and P. Moreno In this paper, we propose a neural search algorithm to select the most likely hypothesis using a sequence of acoustic representations and multiple hypotheses as input.
1568 Transformer Transducer: A Streamable Speech Recognition Model with Transformer Encoders and RNN-T Loss Q. Zhang et al. In this paper we present an end-to-end speech recognition model with Transformer encoders that can be used in a streaming speech recognition system.
1569 Full-Sum Decoding for Hybrid Hmm Based Speech Recognition Using LSTM Language Model W. Zhou, R. Schl?ter and H. Ney We explore the potential gain from more accurate probabilities in terms of decision making and apply the full-sum decoding with a modified prefix-tree search framework.
1570 The Rwth Asr System for Ted-Lium Release 2: Improving Hybrid Hmm With Specaugment W. Zhou, W. Michel, K. Irie, M. Kitza, R. Schl?ter and H. Ney We present a complete training pipeline to build a state-of-the-art hybrid HMM-based ASR system on the 2nd release of the TED-LIUM corpus.
1571 Meta Learning for End-To-End Low-Resource Speech Recognition J. Hsu, Y. Chen and H. Lee In this paper, we proposed to apply meta learning approach for low-resource automatic speech recognition (ASR).
1572 Cross-Speaker Silent-Speech Command Word Recognition Using Electro-Optical Stomatography S. Stone and P. Birkholz In this work, we present the results of a study using a novel measurement technology called Electro-Optical Stomatography to capture speech movements and use the acquired data to recognize a number of command words.
1573 Exploring A Zero-Order Direct Hmm Based on Latent Attention for Automatic Speech Recognition P. Bahar, N. Makarov, A. Zeyer, R. Schl?ter and H. Ney In this paper, we study a simple yet elegant latent variable attention model for automatic speech recognition (ASR) which enables an integration of attention sequence modeling into the direct hidden Markov model (HMM) concept.
1574 Improving Device Directedness Classification of Utterances With Semantic Lexical Features K. Gillespie, I. C. Konstantakopoulos, X. Guo, V. T. Vasudevan and A. Sethy We propose a directedness classifier that combines semantic lexical features with a lightweight acoustic feature and show it is effective in classifying directedness.
1575 Training ASR Models By Generation of Contextual Information K. Singh et al. In this paper, we conduct a large-scale study evaluating the effectiveness of weakly-supervised learning for speech recognition by using loosely related contextual information as a surrogate for ground-truth labels.
1576 Speech Recognition Model Compression M. Sakthi, A. Tewfik and R. Pawate In this paper, we propose Bin & Quant (B&Q), a compression technique using which we were able to reduce the Deep Speech 2 speech recognition model size by 7 times for a negligible loss in accuracy.
1577 Gpu-Accelerated Viterbi Exact Lattice Decoder for Batched Online and Offline Speech Recognition H. Braun, J. Luitjens, R. Leary, T. Kaldewey and D. Povey We present an optimized weighted finite-state transducer (WFST) decoder capable of online streaming and offline batch processing of audio using Graphics Processing Units (GPUs).
1578 Sequence-to-Sequence Automatic Speech Recognition with Word Embedding Regularization and Fused Decoding A. H. Liu, T. Sung, S. Chuang, H. Lee and L. Lee In this paper, we investigate the benefit that off-the-shelf word embedding can bring to the sequence-to-sequence (seq-to-seq) automatic speech recognition (ASR).
1579 Synchronous Transformers for end-to-end Speech Recognition Z. Tian, J. Yi, Y. Bai, J. Tao, S. Zhang and Z. Wen In this paper, we propose a model named synchronous transformer to address this problem, which can predict the output sequence chunk by chunk.
1580 Investigation of Methods to Improve the Recognition Performance of Tamil-English Code-Switched Data in Transformer Framework M. S. Mary N J, V. M. Shetty and S. Umesh In this paper, we investigate methods to deal with such very low resource scenarios.
1581 Bangla Voice Command Recognition in end-to-end System Using Topic Modeling based Contextual Rescoring N. Sadeq, S. Ahmed, S. S. Shubha, M. N. Islam and M. A. Adnan In this work, we perform contextual rescoring using multi-label topic modeling to improve the performance of an End-to-End Bangla voice command recognition system.
1582 Learning to Detect Keyword Parts and Whole by Smoothed Max Pooling H. Park, P. Violette and N. Subrahmanya We propose smoothed max pooling loss and its application to keyword spotting systems.
1583 End-end Speech-to-Text Translation with Modality Agnostic Meta-Learning S. Indurthi et al. In this work, we adopt a meta-learning algorithm to train a modality agnostic multi-task model that transfers knowledge from source tasks=ASR+MT to target task=ST where the ST task severely lacks data.
1584 Analyzing ASR Pretraining for Low-Resource Speech-to-Text Translation M. C. Stoian, S. Bansal and S. Goldwater Here, we experiment with pretraining on datasets of varying sizes, including languages related and unrelated to the AST source language.
1585 Instance-based Model Adaptation for Direct Speech Translation M. A. Di Gangi, V. Nguyen, M. Negri and M. Turchi We tackle this limitation with a method to improve data exploitation and boost the system’s performance at inference time.
1586 Re-Translation Strategies for Long Form, Simultaneous, Spoken Language Translation N. Arivazhagan, C. Cherry, I. Te, W. Macherey, P. Baljekar and G. Foster We investigate the problem of simultaneous machine translation of long-form speech content.
1587 Skinaugment: Auto-Encoding Speaker Conversions for Automatic Speech Translation A. D. McCarthy, L. Puzon and J. Pino We propose autoencoding speaker conversion for training data augmentation in automatic speech translation.
1588 End-to-End Speech Translation with Self-Contained Vocabulary Manipulation M. Tu, F. Zhang and W. Liu But vocabulary manipulation is hard to apply to the end-to-end speech-text translation, because neither source text nor speech-to-target mapping is available. We introduce a method that avoids this dependence.
1589 An Empirical Study of Transformer-Based Neural Language Model Adaptation K. Li et al. We explore two adaptation approaches of deep Transformer based neural language models (LMs) for automatic speech recognition.
1590 Low-bit Quantization of Recurrent Neural Network Language Models Using Alternating Direction Methods of Multipliers J. Xu, X. Chen, S. Hu, J. Yu, X. Liu and H. Meng By formulating quantized RNNLMs training as an optimization problem, this paper presents a novel method to train quantized RNNLMs from scratch using alternating direction methods of multipliers (ADMM).
1591 Audio-Attention Discriminative Language Model for ASR Rescoring A. Gandhe and A. Rastrow In this work, we propose to combine the benefits of end-to-end approaches with a conventional system using an attention-based discriminative language model that learns to rescore the output of a first-pass ASR system.
1592 Training Code-Switching Language Model with Monolingual Data S. Chuang, T. Sung and H. Lee In this paper, we propose an approach to train code-switching language models with monolingual data only.
1593 Domain Robust, Fast, and Compact Neural Language Models A. Gerstenberger, K. Irie, P. Golik, E. Beck and H. Ney We propose training methods for building neural language models for such a task, which are not only domain robust, but reasonable in model size and fast for evaluation.
1594 A Random Gossip BMUF Process for Neural Language Modeling Y. Huang et al. In this paper, we present a decentralized BMUF process, in which the model is split into different components, each of which is updated by communicating to some randomly chosen neighbor nodes with the same component, followed by a BMUF-like process.
1595 Pseudo Labeling and Negative Feedback Learning for Large-Scale Multi-Label Domain Classification J. Kim and Y. Kim In this paper, given one ground-truth domain for each training utterance, we regard domains consistently predicted with the highest confidences as additional pseudo labels for the training.
1596 Pre-Training for Query Rewriting in a Spoken Language Understanding System Z. Chen, X. Fan and Y. Ling In this work, we first propose a neural-retrieval based approach for query rewriting. Then, inspired by the wide success of pre-trained contextual language embeddings, and also as a way to compensate for insufficient QR training data, we propose a language-modeling (LM) based approach to pre-train query embeddings on historical user conversation data with a voice assistant.
1597 End-to-End Architectures for ASR-Free Spoken Language Understanding E. Palogiannidi, I. Gkinis, G. Mastrapas, P. Mizera and T. Stafylakis In this paper, we explore a set of recurrent architectures for intent classification, tailored to the recently introduced Fluent Speech Commands (FSC) dataset, where intents are formed as combinations of three slots (action, object, and location).
1598 End-To-End Spoken Language Understanding Without Matched Language Speech Model Pretraining Data R. Price This paper proposes two strategies to improve the performance of E2E SLU models in scenarios where transcribed data for pretraining in the target language is unavailable: multilingual pretraining with mismatched languages and data augmentation using SpecAugment[1].
1599 Leveraging Unpaired Text Data for Training End-To-End Speech-to-Intent Systems Y. Huang et al. In this paper, we attempt to leverage NLU text resources.
1600 Generating Empathetic Responses by Looking Ahead the User?s Sentiment J. Shin, P. Xu, A. Madotto and P. Fung Hence, in this paper, we propose Sentiment Look-ahead, which is a novel perspective for empathy that models the future user emotional state.
1601 A Hierarchical Model for Dialog Act Recognition Considering Acoustic and Lexical Context Information Y. Si, L. Wang, J. Dang, M. Wu and A. Li To solve the problem, we propose a hierarchical model for DAR considering context information of both lexical and acoustic prosody.
1602 Large-Scale Unsupervised Pre-Training for End-to-End Spoken Language Understanding P. Wang, L. Wei, Y. Cao, J. Xie and Z. Nie In this paper, we explore unsupervised pre-training for End-to-end SLU models by learning representations from large-scale raw audios.
1603 Improving Spoken Question Answering Using Contextualized Word Representation D. Su and P. Fung In this paper, we propose using contextualized word representations to mitigate the effects of ASR errors and pretraining on existing textual QA datasets to mitigate the data scarcity issue.
1604 Learning Asr-Robust Contextualized Embeddings for Spoken Language Understanding C. Huang and Y. Chen We propose a novel confusion-aware fine-tuning method to mitigate the impact of ASR errors on pre-trained LMs.
1605 A Hierarchical Tracker for Multi-Domain Dialogue State Tracking J. Li, S. Zhu and K. Yu In this paper, we propose a hierarchical dialogue state tracker which consists of three sequential modules: domain classification, slot detection and value extraction.
1606 A BI-Model Approach for Handling Unknown Slot Values in Dialogue State Tracking Y. Wang, Y. Shen and H. Jin In this paper, we present an end-to-end bi-model structure for dialogue state tracking, which can handle the scenarios when the spoken language understanding model with a predefined slot candidate list is absent.
1607 Improving Sample-Efficiency in Reinforcement Learning for Dialogue Systems by Using Trainable-Action-Mask Y. Wu, B. Tseng and C. E. Rasmussen In this paper, we propose trainable-action-mask (TAM) which learns from data automatically without handcrafting complicated rules.
1608 Fg2seq: Effectively Encoding Knowledge for End-To-End Task-Oriented Dialog Z. He, Y. He, Q. Wu and J. Chen In this paper, we propose a Flow-to-Graph seq2seq model (FG2Seq) which can effectively encode knowledge by considering inherent structural information of the knowledge graph and latent semantic information from dialog history.
1609 A Simple But Effective Bert Model for Dialog State Tracking on Resource-Limited Systems T. M. Lai, Q. Hung Tran, T. Bui and D. Kihara In this work, we propose a simple but effective DST model based on BERT.
1610 Fast Domain Adaptation for Goal-Oriented Dialogue Using a Hybrid Generative-Retrieval Transformer I. Shalyminov, A. Sordoni, A. Atkinson and H. Schulz In this paper, we present a hybrid generative-retrieval model that can be trained using transfer learning.
1611 Predicting Performance Outcome with a Conversational Graph Convolutional Network for Small Group Interactions Y. Lin and C. Lee By introducing the use of the graph structure in modeling the natural inter-member conversational ties during such an interaction, we aim to advance the state-of-art computational approach in predicting group performance scores.
1612 Design Considerations for Hypothesis Rejection Modules in Spoken Language Understanding Systems A. Alok, R. Gupta and S. Ananthakrishnan In this work, we present two designs for SLU hypothesis rejection modules: (i) scheme R1 that performs rejection on domain specific SLU hypothesis and, (ii) scheme R2 that performs rejection on hypothesis generated from the overall SLU system.
1613 Combining Acoustics, Content and Interaction Features to Find Hot Spots in Meetings D. Makhervaks, W. Hinthorn, D. Dimitriadis and A. Stolcke In this paper we investigate to what extent various acoustic, linguistic and pragmatic aspects of the meetings, both in isolation and jointly, can help detect hot spots.
1614 Accent Estimation of Japanese Words from Their Surfaces and Romanizations for Building Large Vocabulary Accent Dictionaries H. Tachibana and Y. Katayama In order to build a large scale accent dictionary that contains those words, the authors developed an accent estimation technique that predicts the accent of a word from its limited information, namely the surface (e.g. kanji) and the yomi (simplified phonetic information).
1615 OH, JEEZ! or UH-HUH? A Listener-Aware Backchannel Predictor on ASR Transcriptions D. Ortega, C. Li and N. T. Vu This paper presents our latest investigation on modeling backchannel in conversations.
1616 Controllable Time-Delay Transformer for Real-Time Punctuation Prediction and Disfluency Detection Q. Chen, M. Chen, B. Li and W. Wang In this paper, we propose a Controllable Time-delay Transformer (CT-Transformer) model that jointly completes the punctuation prediction and disfluency detection tasks in real time.
1617 Identifying Truthful Language in Child Interviews V. Ardulov, Z. Durante, S. Williams, T. Lyon and S. Narayanan The work presented uses various machine learning algorithms to identify differences in the speech of children when they are lying or being truthful, particularly when they have been asked by a confederate to deceive an interviewer.
1618 A multi-view approach for Mandarin non-native mispronunciation verification Z. Wana, J. H. L. Hansen and Y. Xie In this study, a multi-view approach is proposed to incorporate discriminative feature representations which requires less annotation for non-native mispronunciation verification of Mandarin.
1619 Diacritic-Level Pronunciation Analysis Using Phonological Features A. Kain, A. Roten and R. Gale We propose a system that combines a multi-target architecture with weighted finite-state transducers to first segment and then analyze an utterance in terms of its phonological features.
1620 Phoneme Boundary Detection Using Learnable Segmental Features F. Kreuk, Y. Sheena, J. Keshet and Y. Adi In this work, we propose a neural architecture coupled with a parameterized structured loss function to learn segmental representations for the task of phoneme boundary detection.
1621 Meta-Learning for Robust Child-Adult Classification from Speech N. R. Koluguri, M. Kumar, S. H. Kim, C. Lord and S. Narayanan In this paper, we address a specific sub-problem of speaker diarization, namely child-adult speaker classification in such dyadic conversations with specified roles.
1622 Accounting for Microprosody in Modeling Intonation P. Birkholz and X. Zhang Here, we derived models for two forms of microvariations: the drop in f0 during voiced obstruents, and the increased f0 at the onset of vowels following voiceless obstruents.
1623 How confident are you? Exploring the role of fillers in the automatic prediction of a speaker?s confidence T. Dinkar, I. Vasilescu, C. Pelachaud and C. Clavel We introduce a new and challenging task, that is the prediction of FOAK, which we think has widespread applicability, given the increasing popularity of automatic processing of educational and job interviews, reviews and speeches.
1624 Training Spoken Language Understanding Systems with Non-Parallel Speech and Text L. Sari, S. Thomas and M. Hasegawa-Johnson In this study, we investigate the use of non-parallel speech and text to improve the performance of dialog act recognition as an example SLU task.
1625 What is best for spoken language understanding: small but task-dependant embeddings or huge but out-of-domain embeddings? S. Ghannay, A. Neuraz and S. Rosset The goal of this study is two-fold: firstly, it focuses on semantic evaluation of common word embeddings approaches for SLU task; secondly, it investigates the use of two different data sets to train the embeddings: small and task-dependent corpus or huge and out-of-domain corpus.
1626 Fast Intent Classification for Spoken Language Understanding Systems A. Tyagi et al. To address the latency and computational complexity issues, we explore a BranchyNet scheme on an intent classification scheme within SLU systems.
1627 Converting Written Language to Spoken Language with Neural Machine Translation for Language Modeling S. Ando, M. Suzuki, N. Itoh, G. Kurata and N. Minematsu We address this problem by generating texts in spoken language from those in written language by using a neural machine translation (NMT) model.
1628 Addressing the Polysemy Problem in Language Modeling with Attentional Multi-Sense Embeddings R. Ma, L. Jin, Q. Liu, L. Chen and K. Yu In this work, we address this problem by assigning multiple fine-grained sense embeddings to each word in the embedding layers.
1629 Addressing Challenges in Building Web-Scale Content Classification Systems A. S. Timmaraju, A. Liu and P. Tripathi We present learnings from our efforts in building a content classification system for multiple document types at Facebook using Multi-modal Transformers.
1630 A Neural Document Language Modeling Framework for Spoken Document Retrieval L. Yen, Z. Wu and K. Chen Targeting on enhancing the SDR performance, the paper concentrates on proposing a neural retrieval framework, which assembles the merits of using language modeling (LM) mechanism in SDR and leveraging the abstractive information learned by the language representation models.
1631 Spoken Document Retrieval Leveraging Bert-Based Modeling and Query Reformulation S. Fan-Jiang, T. Lo and B. Chen In this paper, we present a novel study of SDR leveraging the Bidirectional Encoder Representations from Transformers (BERT) model for query and document representations (embeddings), as well as for relevance scoring.
1632 Multitask Learning for Darpa Lorelei?s Situation Frame Extraction Task K. Singla and S. Narayanan This paper describes a novel approach of multitask learning for an end-to-end optimization technique for document classification.
1633 Auxiliary Capsules for Natural Language Understanding I. Staliunaite and I. Iacobacci In this work we extend the newly introduced application of Capsule Networks for NLU to a multi-task learning environment, using relevant auxiliary tasks.
1634 Discrete Wasserstein Autoencoders for Document Retrieval Y. Zhang and H. Zhu In this paper, we present an end-to-end Wasserstein Autoencoder (WAE) for text hashing to avoid in-differentiable operators in the reparameterization trick, where the latent variables can be imposed to any discrete priors we can sample by using adversarial learning.
1635 Cross-Lingual Topic Prediction For Speech Using Translations S. Bansal, H. Kamper, A. Lopez and S. Goldwater Given a large amount of unannotated speech in a low-resource language, can we classify the speech utterances by topic? We consider this question in the setting where a small amount of speech in the low-resource language is paired with text translations in a high-resource language.
1636 Embedded Large-Scale Handwritten Chinese Character Recognition Y. Chherawala, H. J. G. A. Dolfing, R. S. Dixon and J. R. Bellegarda This paper describes how the Apple deep learning recognition system can accurately handle up to 30,000 Chinese characters while running in real-time across a range of mobile devices.
1637 Leveraging Gans to Improve Continuous Path Keyboard Input Models A. Mehra, J. R. Bellegarda, O. Bapat, P. Lal and X. Wang In this work, we address this challenge by using GANs to augment our training corpus with user-realistic synthetic data.
1638 Unsupervised Key Hand Shape Discovery of Sign Language Videos with Correspondence Sparse Autoencoders R. D. Siyli, B. Gundogdu, M. Saraclar and L. Akarun In this paper, we assign labels of an isolated Sign Language(SL) dataset using end-to-end neural network architectures that have proven success in unsupervised discovery of sub-word acoustic units in speech processing.
1639 Keyword Search for Sign Language N. C. Tamer and M. Sara?lar In this paper, we propose a keyword search (KWS) system for a sign language.
1640 Large-Context Pointer-Generator Networks for Spoken-to-Written Style Conversion M. Ihori, A. Takashima and R. Masumura This paper introduces a spoken-to-written style conversion method that is suitable for handling a series of text such as discourses and conversations.
1641 From Unsupervised Machine Translation to Adversarial Text Generation A. Rashid, A. Do-Omri, M. A. Haidar, Q. Liu and M. Rezagholizadeh We present a self-attention based bilingual adversarial text generator (B-GAN) which can learn to generate text from the encoder representation of an unsupervised neural machine translation system.
1642 Self-Attention and Retrieval Enhanced Neural Networks for Essay Generation W. Wang, H. Zheng and Z. Lin In this paper, we focus on essay generation, which aims at generating an essay (a paragraph) according to a set of topic words.
1643 Bert is Not All You Need for Commonsense Inference S. Park, J. Son, S. Hwang and K. Park Our contribution is observing the limitations of BERT in commonsense inference, then leveraging complementary resources containing missing information.
1644 Semi-Supervised Sentence Classification Based on User Polarity in the Social Scenarios B. Ma, H. Sun, J. Wang, Q. Qi and J. Liao Thus, in this paper, a concept called user polarity is proposed to quantify the tendency of sentences published by a user which are categorized into the same class.
1645 Upgrading CRFS to JRFS and its Benefits to Sequence Modeling and Labeling Y. Song, Z. Ou, Z. Liu and S. Yang In this paper, we propose to upgrade condtional random fields (CRFs) and obtain a joint generative model of observation and label sequences, called joint random fields (JRFs).
1646 Selective Attention Encoders by Syntactic Graph Convolutional Networks for Document Summarization H. Xu, Y. Wang, K. Han, B. Ma, J. Chen and X. Li In this paper, we propose to use a graph to connect the parsing trees from the sentences in a document and utilize the stacked graph convolutional networks (GCNs) to learn the syntactic representation for a document.
1647 Learning to Generate Diverse Questions from Keywords Y. Pan, B. Hu, Q. Chen, Y. Xiang and X. Wang In this paper, we focus on a more complex question generation task, i.e., generating a series of questions for each set of keywords (one-to-many). To evaluate the effectiveness of the proposed model, we collect a dataset which contains 62835 questions with respect to 12567 sets of keywords.
1648 Europarl-ST: A Multilingual Corpus for Speech Translation of Parliamentary Debates J. Iranzo-S?nchez et al. In this paper we present Europarl-ST, a novel multilingual SLT corpus containing paired audio-text samples for SLT from and into 6 European languages, for a total of 30 different translation directions.
1649 Multilingual Grapheme-To-Phoneme Conversion with Byte Representation M. Yu et al. In this work, we propose a multilingual G2P model with byte-level input representation to accommodate different grapheme systems, along with an attention-based Transformer architecture.
1650 Language-Agnostic Multilingual Modeling A. Datta, B. Ramabhadran, J. Emond, A. Kannan and B. Roark In this paper, we propose a new approach to building a language-agnostic multilingual ASR system which transforms all languages to one writing system through a many-to-one transliteration transducer.
1651 ADI17: A Fine-Grained Arabic Dialect Identification Dataset S. Shon, A. Ali, Y. Samih, H. Mubarak and J. Glass In this paper, we describe a method to collect dialectal speech from YouTube videos to create a large-scale Dialect Identification (DID) dataset.
1652 Universal Phone Recognition with a Multilingual Allophone System X. Li et al. In this work, we propose a joint model of both language-independent phone and language-dependent phoneme distributions.
1653 Coupled Training of Sequence-to-Sequence Models for Accented Speech Recognition V. Unni, N. Joshi and P. Jyothi We propose coupled training for encoder-decoder ASR models that acts on pairs of utterances corresponding to the same text spoken by speakers with different accents.
1654 Addressing Accent Mismatch In Mandarin-English Code-Switching Speech Recognition Z. Tan, X. Fan, H. Zhu and E. Lin In this paper, we experiment on Mandarin-English code-switching audio spoken by native Chinese speakers and evaluate three techniques to improve accuracy – data adaptation, individual senone modeling and lexicon enrichment.
1655 Detecting Mismatch Between Text Script and Voice-Over Using Utterance Verification Based on Phoneme Recognition Ranking Y. Jeong and H. Cho The purpose of this study is to detect the mismatch between text script and voice-over.
1656 DNN-Based Speech Recognition for Globalphone Languages M. Y. Tachbelie, A. Abulimiti, S. T. Abate and T. Schultz This paper describes new reference benchmark results based on hybrid Hidden Markov Model and Deep Neural Networks (HMM-DNN) for the GlobalPhone (GP) multilingual text and speech database.
1657 Deep Neural Networks Based Automatic Speech Recognition for Four Ethiopian Languages S. T. Abate, M. Yifiru Tachbelie and T. Schultz In this work, we present speech recognition systems for four Ethiopian languages: Amharic, Tigrigna, Oromo and Wolaytta.
1658 Improving the Performance of Transformer Based Low Resource Speech Recognition for Indian Languages V. M. Shetty, M. Sagaya Mary N J and S. Umesh In this work, we explore the application of Transformers on low resource Indian languages in a multilingual framework.
1659 Improving Language Identification for Multilingual Speakers A. Titus, J. Silovsky, N. Chen, R. Hsiao, M. Young and A. Ghoshal We address this by using coarser-grained targets for the acoustic LID model and integrating its outputs with interaction context signals in a context-aware model to tailor the system to each user.
1660 Information Flow Optimization in Inference Networks A. Deshmukh, J. Liu, V. V. Veeravalli and G. Verma The problem of maximizing the information flow through a sensor network tasked with an inference objective at the fusion center is considered.
1661 Exploiting Two-Dimensional Symmetry and Unimodality for Model-Free Source Localization in Harsh Environment J. Chen This paper studies a model-free source localization problem where sensors only measure received signal strengths from a wireless signal source.
1662 Uncertainty Quantification for Remaining Useful Lifetime Prediction with Multi-Channel Sensory Data Y. Deng, H. Wu and C. Jiang In this paper, the ratio of mean to variance was considered to measure the uncertainty propagation rate (UPR) of RUL prediction over time.
1663 ViMo: Vital Sign Monitoring Using Commodity Millimeter Wave Radio F. Wang, F. Zhang, C. Wu, B. Wang and K. J. Ray Liu In this paper, we propose ViMo, a calibration-free remote Vital sign Monitoring system that can simultaneously monitor multiple users by leveraging the channel impulse response (CIR) of 60GHz WiFi.
1664 Time Reversal Based Robust Gesture Recognition Using Wifi S. D. Regani, B. Wang, M. Wu and K. J. Ray Liu In this work, we propose WiGRep, a time reversal based gesture recognition approach using Wi-Fi, which can recognize different gestures by counting the number of repeating gesture segments.
1665 Noncoherent Maximum-Likelihood Detection for Ambient Backscattering Communications Over Ambient OFDM Signals D. Darsena This paper deals with the problem of noncoherent maximum-likelihood (ML) signal detection for backscatter communications over ambient OFDM.
1666 Bandit Sampling for Faster Activity and Data Detection in Massive Random Access J. Dong, J. Zhang and Y. Shi This paper considers the grant-free random access scheme in IoT networks with a massive number of devices.
1667 Consensus-Based Distributed Clustering for IoT H. Chen, H. Yu, S. Zhao and Q. Shi Facing the distributed network challenges including data volume, communication latency, and information security, we here propose a distributed clustering algorithm where each IoT device may have data from multiple clusters.
1668 Energy-Efficient 3D UAV Trajectory Design for Data Collection in Wireless Sensor Networks D. B. Licea, E. Nurellari and M. Ghogho We consider the issue of designing closed 3D UAV trajectories that allow for an energy efficient collection of data with a UAV-aided wireless sensor network.
1669 Application Informed Motion Signal Processing for Finger Motion Tracking Using Wearable Sensors Y. Liu, F. Jiang and M. Gowda This paper presents a system called FinGTrAC that shows the feasibility of fine grained finger gesture tracking using low intrusive wearable sensor platform (smart-ring worn on the index finger and a smart-watch worn on the wrist).
1670 Instant Adaptive Learning: An Adaptive Filter Based Fast Learning Model Construction for Sensor Signal Time Series Classification on Edge Devices A. Pal, A. Ukil, T. Deb, I. Sahu and A. Majumdar In this paper, we propose Instant Adaptive Learning that characterizes the intrinsic signal processing properties of time series sensor signals using linear adaptive filtering and derivative spectrum to efficiently construct a low-cost learning model followed by standard classification algorithms.
1671 Image Fusion using Joint Sparse Representations and Coupled Dictionary Learning F. G. Veshki, N. Ouzir and S. A. Vorobyov In this work, to identify the sharp image patches, we propose an improved discriminative coupled dictionary learning approach using joint sparse representations in blurred and focused dictionaries.
1672 Clustering of Nonnegative Data and an Application to Matrix Completion C. Strohmeier and D. Needell In this article, we propose a simple algorithm to cluster nonnegative data lying in disjoint subspaces.
1673 Sampling of Surfaces and Learning Functions in High Dimensions Q. Zou and M. Jacob To capture the non-linear structure of the data, we model the data as points living on a smooth surface. We model the surface as the zero level-set of a bandlimited function.
1674 Stock Movement Prediction That Integrates Heterogeneous Data Sources Using Dilated Causal Convolution Networks with Attention D. Daiya, M. Wu and C. Lin The purpose of this research is to develop a high performing model for stock movement prediction utilizing financial indicators and news data.
1675 Large-Scale Weakly-Supervised Content Embeddings for Music Recommendation and Tagging Q. Huang, A. Jansen, L. Zhang, D. P. W. Ellis, R. A. Saurous and J. Anderson We explore content-based representation learning strategies tailored for large-scale, uncurated music collections that afford only weak supervision through unstructured natural language metadata and colisten statistics.
1676 Supervised Canonical Correlation Analysis of Data on Symmetric Positive Definite Manifolds by Riemannian Dimensionality Reduction F. Fallah and B. Yang In this paper, we propose a framework for a supervised CCA of manifold-based data.
1677 The Empirical Duality Gap of Constrained Statistical Learning L. F. O. Chamon, S. Paternain, M. Calvo-Fullana and A. Ribeiro In this work, we propose to directly tackle the constrained statistical problem overcoming its infinite dimensionality, unknown distributions, and constraints by leveraging finite dimensional parameterizations, sample averages, and duality theory.
1678 Context and Uncertainty Modeling for Online Speaker Change Detection H. Aronowitz and W. Zhu In this work we focus on online speaker change detection as a standalone task which is required for online closed captioning of broadcast television.
1679 Modeling Uncertainty in Predicting Emotional Attributes from Spontaneous Speech K. Sridhar and C. Busso An intriguing approach to predict uncertainty is Monte Carlo (MC) dropout, which obtains predictions from multiple feed-forward passes through a deep neural network (DNN) by using dropout regularization in both training and inference. This study evaluates this approach with regression models to predict emotional attribute scores for valence, arousal and dominance.
1680 Accuracy-Robustness Trade-Off for Positively Weighted Neural Networks A. Neacsu, J. Pesquet and C. Burileanu We propose a stochastic projected gradient descent algorithm which allows us to adjust this constant in the training process.
1681 Towards a New Understanding of the Training of Neural Networks with Mislabeled Training Data H. Gish, J. Silovsky, M. Sung, M. Siu, W. Hartmann and Z. Jiang We investigate the problem of machine learning with mislabeled training data.
1682 On Network Science and Mutual Information for Explaining Deep Neural Networks B. Davis, U. Bhatt, K. Bhardwaj, R. Marculescu and J. M. F. Moura In this paper, we present a new approach to interpret deep learning models.
1683 Spatial Active Noise Control Based on Kernel Interpolation with Directional Weighting H. Ito, S. Koyama, N. Ueno and H. Saruwatari A spatial active noise control (ANC) method taking prior information on the approximate direction of primary noise sources into consideration is proposed.
1684 Active Noise Control Over Multiple Regions: Performance Analysis J. Zhang, H. Sun, P. N. Samarasinghe and T. D. Abhayapala In this paper, we perform an initial study on the more complex problem of simultaneous noise control over multiple target regions using a single ANC system.
1685 Array-Geometry-Aware Spatial Active Noise Control Based on Direction-of-Arrival Weighting Y. Maeno, Y. Takida, N. Murata and Y. Mitsufuji In this paper, we propose a direction of arrival (DOA) weighting algorithm for the adaptive filter update, which prioritizes residual error control with respect to the array geometry.
1686 Multichannel Active Noise Control with Spatial Derivative Constraints to Enlarge the Quiet Zone D. Shi, B. Lam, S. Wen and W. Gan This paper proposes a time-domain adaptive algorithm to implement the spatial derivative constraint, which avoids the complex analytic acoustic calculations.
1687 An Acoustic Modelling Based Remote Error Sensing Approach for Quiet Zone Generation in a Noisy Environment Q. Zhu, X. Qiu and I. Burnett This paper presents an improved approach to increase the effective frequency range based on the acoustic modelling.
1688 Distributed Wave-Domain Active Noise Control Based on the Diffusion Strategy Y. Dong, J. Chen and W. Zhang In order to address this issue, this work presents a distributed wave-domain ANC scheme by resorting to distributed optimization techniques.
1689 A Theoretical Basis for Practitioners Heuristic 1/N and Long-Only Quintile Portfolio R. Zhou and D. P. Palomar In this paper, we formulate a mathematically meaningful robust maximum return portfolio design and show that it reduces to the two heuristic portfolios under different level of estimation error in the mean returns.
1690 A Recursive Bayesian Solution for the Excess Over Threshold Distribution with Stochastic Parameters D. E. Johnston and P. M. Djuric In this paper, we propose a new approach for analyzing extreme values that are witnessed in financial markets.
1691 Gaussian Process Imputation of Multiple Financial Series T. d. Wolff, A. Cuevas and F. Tobar We focus on learning the relationships among financial time series by modelling them through a multi-output Gaussian process (MOGP) with expressive covariance functions.
1692 Robust Covariance Matrix Estimation and Portfolio Allocation: The Case of Non-Homogeneous Assets E. Jay, T. Soler, J. -. Ovarlez, P. D. Peretti and C. Chorro This paper presents how the most recent improvements made on covariance matrix estimation and model order selection can be applied to the portfolio optimization problem.
1693 Portfolio Cuts: A Graph-Theoretic Framework to Diversification B. S. Dees, L. Stankovic, A. G. Constantinides and D. P. Mandic To this end, we investigate ways for domain knowledge to be conveniently incorporated into the analysis, by means of graphs.
1694 CORRGAN: Sampling Realistic Financial Correlation Matrices Using Generative Adversarial Networks G. Marti We propose a novel approach for sampling realistic financial correlation matrices.
1695 Graph Neural Net Using Analytical Graph Filters and Topology Optimization for Image Denoising W. Su, G. Cheung, R. Wildes and C. Lin Inspired by an analytically derived CNN by Hadji et al., in this paper we construct a new layered graph neural net (GNN) using GraphBio as our graph filter.
1696 Defending Graph Convolutional Networks Against Adversarial Attacks V. N. Ioannidis and G. B. Giannakis This paper introduces graph neural network architectures that are robust to perturbed networked data.
1697 Constrained Spectral Clustering for Dynamic Community Detection A. Karaaslanli and S. Aviyente In this paper, we propose a new stochastic block model (SBM) for modeling the evolution of community membership.
1698 Towards an Efficient and General Framework of Robust Training for Graph Neural Networks K. Xu et al. To overcome these limitations, we propose a general framework which leverages the greedy search algorithms and zeroth-order methods to obtain robust GNNs in a generic and an efficient manner.
1699 Deep Geometric Knowledge Distillation with Graphs C. Lassance, M. Bontonou, G. B. Hacene, V. Gripon, J. Tang and A. Ortega In this work, we focus instead on relative knowledge distillation (RKD), which considers the geometry of the respective latent spaces, allowing for dimension-agnostic transfer of knowledge.
1700 On The Choice of Graph Neural Network Architectures C. Vignac, G. Ortiz-Jim?nez and P. Frossard In this paper, we show empirically that in settings with fewer features and more training data, more complex graph networks significantly outperform simple models, and propose a few insights towards the proper choice of graph network architectures.
1701 Multitask Learning with Capsule Networks for Speech-to-Intent Applications J. Poncelet and H. V. hamme In this paper, we show how capsules can incorporate multitask learning, which often can improve the performance of a model when the task is difficult.
1702 Using Speech Synthesis to Train End-To-End Spoken Language Understanding Models L. Lugosch, B. H. Meyer, D. Nowrouzezahrai and M. Ravanelli We propose a strategy to overcome this requirement in which speech synthesis is used to generate a large synthetic training dataset from several artificial speakers.
1703 Improved End-To-End Spoken Utterance Classification with a Self-Attention Acoustic Classifier R. Price, M. Mehrabani and S. Bangalore In this paper, we propose a new architecture for end-to-end spoken utterance classification (SUC) and also explore the impact of leveraging lexical information in conjunction with acoustic information obtained from the end-to-end model for SUC.
1704 Dialogue History Integration into End-to-End Signal-to-Concept Spoken Language Understanding Systems N. Tomashenko, C. Raymond, A. Caubri?re, R. D. Mori and Y. Est?ve We proposed to integrate dialogue history into an end-to-end signal-to-concept SLU system.
1705 Error Analysis Applied to End-to-End Spoken Language Understanding A. Caubri?re et al. This paper presents a qualitative study of errors produced by an end-to-end spoken language understanding (SLU) system (speech signal to concepts) that reaches state of the art performance.
1706 A Data Efficient End-to-End Spoken Language Understanding Architecture M. Dinarelli, N. Kapoor, B. Jabaian and L. Besacier In this paper we introduce a data efficient system which is trained end-to-end, with no additional, pre-trained external module.
1707 Federated Neuromorphic Learning of Spiking Neural Networks for Low-Power Edge Intelligence N. Skatchkovsky, H. Jang and O. Simeone In this paper, we propose to mitigate this problem via cooperative training through Federated Learning (FL).
1708 Temporal Coding in Spiking Neural Networks with Alpha Synaptic Function I. M. Comsa, T. Fischbacher, K. Potempa, A. Gesmundo, L. Versari and J. Alakuijala We propose a spiking neural network model that encodes information in the relative timing of individual neuron spikes and performs classification using the first output neuron to spike.
1709 Event-Driven Signal Processing with Neuromorphic Computing Systems P. Blouw and C. Eliasmith In this paper, we provide an overview of tools and methods for building applications that run on neuromorphic computing devices.
1710 Challenges and Perspectives in Neuromorphic-based Visual IoT Systems and Networks M. Martini, N. Khan, Y. Bi, Y. Andreopoulos, H. Saki and M. Shikh-Bahaei Hence, we explore here the feasibility of advanced Machine to Machine (M2M) communications systems that directly capture, compress and transmit spike-based visual information to cloud computing services in order to produce content classification or retrieval results with extremely low power and low latency.
1711 Spiking Neural Networks Trained With Backpropagation for Low Power Neuromorphic Implementation of Voice Activity Detection F. Martinelli, G. Dellaferrera, P. Mainar and M. Cernak We describe a training procedure that achieves low spiking activity and apply pruning algorithms to remove up to 85% of the network connections with no performance loss.
1712 Training Deep Spiking Neural Networks for Energy-Efficient Neuromorphic Computing G. Srinivasan, C. Lee, A. Sengupta, P. Panda, S. S. Sarwar and K. Roy In this paper, we propose different SNN training methodologies, varying in degrees of biofidelity, and evaluate their efficacy on complex image recognition datasets.
1713 Deep Learning for Robust Power Control for Wireless Networks W. Cui, K. Shen and W. Yu This paper aims to show that a deep learning approach for network utility maximization can produce more robust solutions than the traditional model-based approach.
1714 Feedback Turbo Autoencoder Y. Jiang, H. Kim, H. Asnani, S. Oh, S. Kannan and P. Viswanath In this work, we propose Feedback Auto Turbo Encoder (FTAE) which harmoniously combines interleaver and iterative decoding with CNN architectures and demonstrate the blocklength gain and improved performance in the block feedback setting.
1715 Exploiting Channel Locality for Adaptive Massive MIMO Signal Detection M. Khani, M. Alizadeh, J. Hoydis and P. Fleming We propose MMNet, a deep learning MIMO detection scheme that significantly outperforms existing approaches on realistic channels with the same or lower computational complexity.
1716 Deep Learning-Based Beam Alignment in Mmwave Vehicular Networks N. J. Myers, Y. Wang, N. Gonz?lez-Prelcic and R. W. Heath In this paper, we propose an end-to-end deep learning technique to design a structured CS matrix that is well suited to the underlying channel distribution, leveraging both sparsity and the particular spatial structure that appears in vehicular channels.
1717 Joint Source-Channel Coding and Bayesian Message Passing Detection for Grant-Free Radio Access in IoT J. Dommel, Z. Utkovski, S. Stanczak and O. Simeone In contrast, this paper considers a potentially more efficient solution based on Joint Source-Channel (JSC) coding via a non-orthogonal generalization of Type-Based Multiple Access (TBMA).
1718 CNN-Based Analog CSI Feedback in FDD MIMO-OFDM Systems M. B. Mashhadi, Q. Yang and D. G?nd?z Instead, we propose here a Convolutional neural network (CNN)-based analog feedback scheme, called AnalogDeepCMC, which directly maps the downlink CSI to uplink channel input.
1719 Scalable Learning-Based Sampling Optimization for Compressive Dynamic MRI T. Sanchez et al. In this paper, we tackle the problem of designing a sampling mask for an arbitrary reconstruction method and a limited acquisition budget.
1720 Linear Speedup in Saddle-Point Escape for Decentralized Non-Convex Optimization S. Vlaski and A. H. Sayed In this work, we examine in detail the dependence of second-order convergence guarantees on the spectral properties of the combination policy for non-convex multi agent optimization.
1721 On Distributed Stochastic Gradient Algorithms for Global Optimization B. Swenson, A. Sridhar and H. V. Poor The paper considers the problem of network-based computation of global minima in smooth nonconvex optimization problems.
1722 Distributed Tensor Completion Over Networks C. Battiloro and P. D. Lorenzo The aim of this paper is to propose a novel distributed strategy for tensor completion, where (partial) data are collected over a network of agents with sparse, but connected, topology.
1723 Lookahead Converges to Stationary Points of Smooth Non-convex Functions J. Wang, V. Tantia, N. Ballas and M. Rabbat We prove that, with appropriate choice of step-sizes, Lookahead converges to a stationary point of smooth non-convex functions.
1724 Communication Constrained Learning with Uncertain Models J. Z. Hare, C. A. Uribe, L. M. Kaplan and a. Ali Jadbabaie We consider the problem of distributed inference of a group of agents in a social network, where the agents construct, share, and update beliefs in a non-Bayesian framework to identify the underlying true state of the world.
1725 A Sparse Linear Array Approach in Automotive Radars Using Matrix Completion S. Sun and A. P. Petropulu In this paper, different from previous efforts, we use matrix completion to complete the corresponding virtual uniform linear array (ULA) before estimating the target angle.
1726 A Low-Resolution ADC Proof-of-Concept Development for a Fully-Digital Millimeter-wave Joint Communication-Radar P. Kumari, A. Mezghani and R. W. Heath To address these concerns, we present a low-complexity proof-of-concept (PoC) development for a wideband MIMO JCR that uses a mmWave communications waveform and low-resolution ADCs, while maintaining a separate radio-frequency chain per antenna.
1727 Performance Bounds for Displaced Sensor Automotive Radar Imaging G. Wang and K. V. Mishra In this paper, we derive performance bounds on the estimation error of target parameters processed by displaced sensors that correspond to several independent radars mounted at different locations on the same vehicle.
1728 Spatial and Temporal Smoothing for Covariance Estimation in Super-Resolution Angle Estimation in Automotive Radars A. E. Ertan and M. Ali In this paper, we analyze the effect covariance matrix averaging in spatial and temporal domain on the angular resolution that can be obtained with MUSIC.
1729 Slow-Time MIMO-FMCW Automotive Radar Detection with Imperfect Waveform Separation P. Wang, P. Boufounos, H. Mansour and P. V. Orlik We develop an explicit signal model that accounts for waveform separation residuals and propose a Kronecker subspace-based object detector in the framework of generalized likelihood ratio test (GLRT).
1730 On Binary Sequence Set Design with Applications to Automotive Radar R. Lin and J. Li We consider herein the case of two vehicles equipped with multi-input multi-output (MIMO) automotive radars driving next to each other.
1731 Optimal Transport Structure of CycleGAN for Unsupervised Learning for Inverse Problems B. Sim, G. Oh and J. C. Ye In this article, we explain the link between these two framework by mathematical formula and experimental results.
1732 Light-Field Reconstruction and Depth Estimation from Focal Stack Images Using Convolutional Neural Networks Z. Huang, J. A. Fessler, T. B. Norris and I. Y. Chun This paper proposes a non-iterative LF reconstruction and depth estimation method based on three sequential convolutional neural networks (CNNs).
1733 Joint Learning of Cartesian under Sampling Andre Construction for Accelerated MRI T. Weiss, S. Vedula, O. Senouf, O. Michailovich, M. Zibulevsky and A. Bronstein To this end, we propose an algorithm for training the combined acquisition-reconstruction pipeline end-to-end in a differentiable way.
1734 Building Firmly Nonexpansive Convolutional Neural Networks M. Terris, A. Repetti, J. Pesquet and Y. Wiaux In this paper, we develop an optimization algorithm that can be incorporated in the training of a network to ensure the nonexpansiveness of its convolutional layers.
1735 Confirmnet: Convolutional Firmnet and Application to Image Denoising and Inpainting P. K. Pokala, P. Kumar Uttam and C. S. Seelamantula We propose a convolutional iterative firm-thresholding algorithm (CIFTA) building on our previously proposed IFTA, and its deep-unfolded version, namely, convolutional-FirmNet (ConFirmNet).
1736 Learning Differentiable Sparse and Low Rank Networks for Audio-Visual Object Localization J. Pu, Y. Panagakis and M. Pantic In this paper, we propose a novel method to design specific deep neural networks for sparse and low-rank models, where the network can learn a data-adaptive model from training data.
1737 Spherical Large Intelligent Surfaces S. Hu and F. Rusek In this paper, we extend LISs to be three-dimensional (3D) and deployed as spherical surfaces.
1738 Pathloss Prediction using Deep Learning with Applications to Cellular Optimization and Efficient D2D Link Scheduling R. Levie, ?. Yapar, G. Kutyniok and G. Caire In this paper we propose a highly efficient and very accurate method for estimating the propagation pathloss from a point x to all points y on the 2D plane.
1739 Beamforming in Intelligent Environments based on Ultra-Massive MIMO Platforms in Millimeter Wave and Terahertz Bands S. Nie and I. F. Akyildiz In this paper, a joint beamforming scheme is developed based on fractional programming optimization to maximize the spectral efficiency under practical consideration of energy constraints of the UM MIMO communication platform.
1740 A Single-RF Architecture for Multiuser Massive MIMO Via Reflecting Surfaces A. Bereyhi, V. Jamali, R. R. M?ller, A. M. Tulino, G. Fischer and R. Schober In this work, we propose a new single-RF MIMO architecture which enjoys high scalability and energy-efficiency.
1741 Hybrid Precoding for Secure Transmission in Reflect-Array-Assisted Massive MIMO Systems S. Asaad, R. F. Schaefer and H. Vincent Poor Inspired by this architecture, we design a secure multiuser hybrid analog-digital precoding scheme.
1742 Wideband Channel Tracking for Millimeter Wave Massive Mimo Systems with Hybrid Beamforming Reception G. C. Alexandropoulos, E. Vlachos and J. Thompson We present an efficient algorithm for this estimation problem that is based on the alternating direction method of multipliers.
1743 Improving Auditory Attention Decoding Performance of Linear and Non-Linear Methods using State-Space Model A. Aroudi, T. de Taillez and S. Doclo In this paper, we investigate a state-space model using correlation coefficients obtained with a small correlation window to improve the decoding performance of the linear and the non-linear AAD methods.
1744 Towards Decoding Selective Attention from Single-Trial EEG Data in Cochlear Implant users based on Deep Neural Networks W. Nogueira and H. Dolhopiatenko The goal of this work was to investigate the use of non-linear models based on deep neural networks (DNNs) to improve the selective attention decoding accuracy in CI users.
1745 Harmonic/Percussive Sound Separation and Spectral Complexity Reduction of Music Signals for Cochlear Implant Listeners B. Lentz, A. Nagathil, J. Gauer and R. Martin In this work, a music pre-processing scheme is described which combines these approaches and is applicable to a wider variety of music genres.
1746 Bio-Mimetic Attentional Feedback in Music Source Separation A. Bellur and M. Elhilali In this work, we study these competing theories within a deep-network framework for the task of music source separation.
1747 Talker-Independent Speaker Separation in Reverberant Conditions M. Delfarah, Y. Liu and D. Wang To effectively deal with speaker separation and speech dereverberation, we propose a two-stage strategy where reverberant utterances are first separated and then dereverberated.
1748 Evaluation of Joint Auditory Attention Decoding and Adaptive Binaural Beamforming Approach for Hearing Devices with Attention Switching W. Pu, P. Zan, J. Xiao, T. Zhang and Z. Luo In this study, we present the evaluation results of this joint formulation on a new EEG dataset collected on subjects with dynamic switch.
1749 Robust Pricing Mechanism for Resource Sustainability Under Privacy Constraint in Competitive Online Learning Multi-Agent Systems E. Tampubolon and H. Boche Based on the non-cooperative game as the model for agents? interaction and the noisy online mirror ascent as the model for the rationality of the agents, we propose a novel pricing mechanism that gives the agents incentives for sustainable use of the resources.
1750 Neural Network Wiretap Code Design for Multi-Mode Fiber Optical Channels K. Besser, A. Lonnstrom and E. A. Jorswieck In this work, we develop an autoencoder for the multi-mode fiber wiretap channel taking into account the error performance at the legitimate receiver and the information leakage at potential eavesdroppers.
1751 Age-Based Scheduling Policy for Federated Learning in Mobile Edge Networks H. H. Yang, A. Arafa, T. Q. S. Quek and H. Vincent Poor In this paper, based on a metric termed the age of update (AoU), we propose a scheduling policy by jointly accounting for the staleness of the received parameters and the instantaneous channel qualities to improve the running efficiency of FL.
1752 Adversarial Networks for Secure Wireless Communications T. Marchioro, N. Laurenti and D. G?nd?z We propose a data-driven secure wireless communication scheme, in which the goal is to transmit a signal to a legitimate receiver with minimal distortion, while keeping some information about the signal private from an eavesdropping adversary.
1753 Latency-Minimized Design of secure transmissions in UAV-Aided Communications X. Wu, Q. Li, Y. Lu, H. V. Poor, V. C. M. Leung and P. C. Ching In this paper, we propose a latency-minimized transmission scheme for satisfying legitimate users? (LUs?) content requests securely against Eve.
1754 Detect Insider Attacks Using CNN in Decentralized Optimization G. Li, S. X. Wu, S. Zhang and Q. Li This paper studies the security issue of a gossip-based distributed projected gradient (DPG) algorithm, when it is applied for solving a decentralized multi-agent optimization.
1755 Risk Convergence of Centered Kernel Ridge Regression with Large Dimensional Data K. Elkhalil, A. Kammoun, X. Zhang, M. Alouini and T. Y. Al-Naffouri This paper carries out a large dimensional analysis of a variation of kernel ridge regression that we call centered kernel ridge regression (CKRR), also known in the literature as kernel ridge regression with offset.
1756 A Whiteness Test Based on the Spectral Measure of Large Non-Hermitian Random Matrices A. Bose and W. Hachem In the context of multivariate time series, a whiteness test against an MA(1) correlation model is proposed.
1757 On the Limit Distribution of the Canonical Correlation Coefficients Between the Past and the Future of a High-Dimensional White Noise D. Tieplova, P. Loubaton and L. Pastur On the Limit Distribution of the Canonical Correlation Coefficients Between the Past and the Future of a High-Dimensional White Noise
1758 Positive Solutions for Large Random Linear Systems P. Bizeul, M. Clenet and J. Najim We investigate the componentwise positivity of the solution xn depending on the scaling factors, as the dimensions of the system grow to infinity.
1759 On The Frequency Domain Detection of High Dimensional Time Series A. Rosuel, P. Vallet, P. Loubaton and X. Mestre In this paper, we address the problem of detection, in the frequency domain, of a M-dimensional time series modeled as the output of a M ? K MIMO filter driven by a K-dimensional Gaussian white noise, and disturbed by an additive M-dimensional Gaussian colored noise.
1760 Large Dimensional Asymptotics of Multi-Task Learning M. Tiomoko, C. Louart and R. Couillet Based on a random matrix approach, this article proposes an asymptotic analysis of a support vector machine-inspired multitask learning scheme.
1761 Robust Hybrid Beamforming for Satellite-Terrestrial Integrated Networks Z. Lin, M. Lin, B. Champagne, W. Zhu and N. Al-Dhahir In this paper, we propose a novel robust downlink beamforming (BF) design for satellite-terrestrial integrated networks.
1762 In-Network Caching for Hybrid Satellite-Terrestrial Networks Using Deep Reinforcement Learning N. Garg, M. Sellathurai and T. Ratnarajah In this paper, we consider in-network caching where an unavailable content at one BS can be fetched from the nearest BS in the network, before requesting from the content server.
1763 Multigraph Spectral Clustering for Joint Content Delivery and Scheduling in Beam-Free Satellite Communications M. A. Vazquez and A. I. Perez-Neira This paper tackles the problem of user scheduling in satellite content delivery networks with precoding.
1764 Constant-Envelope Precoding for Satellite Systems C. G. Tsinos, A. Arora and B. Ottersten In this paper, Constant-Envelope Precoding techniques are presented for satellite-based communication systems.
1765 Resource Management in the Multibeam NOMA-based Satellite Downlink T. Ram?rez and C. Mosquera Resource Management in the Multibeam NOMA-based Satellite Downlink
1766 Genetic Algorithm Optimized Support Vector Machine in NOMA-based Satellite Networks with Imperfect CSI X. Yan, K. An, C. Wang, W. Zhu, Y. Li and Z. Feng Inspired by the advantages of machine learning (ML) algorithms, we propose an improved support vector machine (SVM) scheme to reduce the inappropriate user pairing risks and enhance the performance of NOMA based satellite networks with imperfect CSI.
1767 Overcoming High Nanopore Basecaller Error Rates for DNA Storage via Basecaller-Decoder Integration and Convolutional Codes S. Chandak et al. We propose a novel approach which overcomes the high error rates in basecalled sequences by integrating a Viterbi error correction decoder with the basecaller, enabling the decoder to exploit the soft information available in the deep learning based basecaller pipeline.
1768 Efficient Constrained Encoders Correcting a Single Nucleotide Edit in DNA Storage K. Cai, X. He, H. M. Kiah and T. Thanh Nguyen In this paper, we investigate codes that corrects a single nucleotide edit and provide linear-time algorithms that encode binary messages into these codes of length n.
1769 Image Processing in DNA C. Pan, S. M. Hossein Tabatabaei Yazdi, S. Kasra Tabatabaei, A. G. Hernandez, C. Schroeder and O. Milenkovic Here we propose the first method for storing quantized images in DNA that uses signal processing and machine learning techniques to deal with error and cost issues without resorting to the use of redundant oligos or rewriting.
1770 Concentration-Based Polynomial Calculations on Nicked DNA T. Chen and M. Riedel In this paper, we introduce a novel scheme for computing polynomial functions on a substrate of nicked DNA.
1771 Capacity of the Erasure Shuffling Channel S. Shin, R. Heckel and I. Shomorony Motivated by DNA-based data storage, we study the erasure shuffling channel.
1772 Achieving the Capacity of the DNA Storage Channel A. Lenz, P. H. Siegel, A. Wachter-Zeh and E. Yaakohi In this paper we analyze storage systems based on these macromolecules from an information theoretic perspective.
1773 Federated Learning with Quantization Constraints N. Shlezinger, M. Chen, Y. C. Eldar, H. V. Poor and S. Cui In this work, we tackle this challenge using tools from quantization theory.
1774 Cooperative Learning VIA Federated Distillation OVER Fading Channels J. Ahn, O. Simeone and J. Kang While prior work studied implementations of FL over wireless fading channels, here we propose wireless protocols for FD and for an enhanced version thereof that leverages an offline communication phase to communicate “mixed-up” covariate vectors.
1775 On the Byzantine Robustness of Clustered Federated Learning F. Sattler, K. M?ller, T. Wiegand and W. Samek In this work we investigate the application of CFL to byzantine settings, where a subset of clients behaves unpredictably or tries to disturb the joint training effort in an directed or undirected way.
1776 Hierarchical Federated Learning ACROSS Heterogeneous Cellular Networks M. S. H. Abad, E. Ozfatura, D. GUndUz and O. Ercetin We consider federated edge learning (FEEL), where mobile users (MUs) collaboratively learn a global model by sharing local updates on the model parameters rather than their datasets, with the help of a mobile base station (MBS).
1777 Overlap Local-SGD: An Algorithmic Approach to Hide Communication Delays in Distributed SGD J. Wang, H. Liang and G. Joshi In this paper, we propose an algorithmic approach named Overlap Local-SGD (and its momentum variant) to overlap communication and computation so as to speedup the distributed training procedure.
1778 Q-GADMM: Quantized Group ADMM for Communication Efficient Decentralized Machine Learning A. Elgabli, J. Park, A. S. Bedi, M. Bennis and V. Aggarwal In this paper, we propose a communication-efficient decen-tralized machine learning (ML) algorithm, coined quantized group ADMM (Q-GADMM).
1779 Deep Soft Interference Cancellation for MIMO Detection N. Shlezinger, R. Fu and Y. C. Eldar In this work we propose a multiuser MIMO receiver which learns to jointly detect in a data-driven fashion, without assuming a specific channel model or requiring CSI.
1780 An Empirical Bayes Approach to Partially Labeled and Shuffled Data Sets A. Dytso and H. V. Poor Specifically, we propose to train model-based empirical Bayes separately on the set of features and the set of labels and combine/mix the two models based on the proportion of unlabeled pairs.
1781 Reinforced Depth-Aware Deep Learning for Single Image Dehazing T. Guo and V. Monga In this work, we proposed a Depth-aware Dehazing using Reinforcement Learning system, denoted as DDRL.
1782 Learning Plug-And-Play Proximal Quasi-Newton Denoisers A. H. Al-Shabili, H. Mansour and P. T. Boufounos We propose a strategy for training a Gaussian denoiser inspired by an unfolded proximal quasi-Newton algorithm, where the noise level of the input signal to the denoiser is estimated in each iteration and at every entry in the signal.
1783 Joint Optimization of Sampling Patterns and Deep Priors for Improved Parallel MRI H. K. Aggarwal and M. Jacob In this work, we investigate the impact of sampling patterns on the quality of the image recovered using the MoDL algorithm.
1784 Learning Sampling and Model-Based Signal Recovery for Compressed Sensing MRI I. A. M. Huijben, B. S. Veeling and R. J. G. van Sloun Realizing that an optimal sampling pattern may depend on the downstream task (e.g. image reconstruction, segmentation, or classification), we here propose joint learning of both task-adaptive k-space sampling and a subsequent model-based proximal-gradient recovery network.
1785 Inferring Dynamic Group Leadership Using Sequential Bayesian Methods Q. Li, S. J. Godsill, J. Liang and B. I. Ahmad This paper presents an online approach for inferring dominant entities in tracked groups from observations.
1786 Scalable Detection and Tracking of Extended Objects F. Meyer and J. L. Williams This paper presents a factor graph formulation and sum-product algorithm (SPA) for scalable detection and tracking of extended objects that generate multiple measurements.
1787 Adversarial Anomaly Detection for Marked Spatio-Temporal Streaming Data S. Zhu, H. S. Yuchi and Y. Xie How to efficiently detect anomalies in these dynamic systems using these streaming event data? This work proposes a novel anomaly detection framework for such event data combining the Long Short-Term Memory (LSTM) and marked spatio-temporal point processes.
1788 Quickest Detection of Growing Dynamic Anomalies in Networks G. Rovatsos, V. V. Veeravalli, D. Towsley and A. Swami The problem of quickest growing dynamic anomaly detection in sensor networks is studied.
1789 Image Segmentation Based Privacy-Preserving Human Action Recognition for Anomaly Detection J. Yan, F. Angelini and S. M. Naqvi In this paper, we prove that human action recognition accuracy mostly depends on contextual data, rather than on privacy-related data.
1790 Prediction oof Vessel Trajectories From AIS Data Via Sequence-To-Sequence Recurrent Neural Networks N. Forti, L. M. Millefiori, P. Braca and P. Willett In this paper, we address the problem of predicting vessel trajectories based on Automatic Identification System (AIS) data.
1791 Near-Optimal Interference Exploitation 1-Bit Massive MIMO Precoding Via Partial Branch-and-Bound A. Li, F. Liu, C. Masouros, Y. Li and B. Vucetic In this paper, we focus on 1-bit precoding for large-scale antenna systems in the downlink based on the concept of constructive interference (CI).
1792 Secure Symbol-Level Miso Precoding Q. Xu, P. Ren and A. L. Swindlehurst In this paper, we modify the technique to exploit the full DI region, but we show that even with this improvement, the general approach is vulnerable to an intelligent eavesdropper who can perform maximum likelihood detection.
1793 Robust Symbol-Level Precoding Via Autoencoder-Based Deep Learning F. Sohrabi, H. V. Cheng and W. Yu This paper proposes an autoencoder-based symbol-level precoding (SLP) scheme for a massive multiple-input multiple-output (MIMO) system operating in a limited-scattering environment.
1794 Constant Envelope Massive MIMO-OFDM Precoding: an Improved Formulation and Solution S. Domouchtsidis, C. Tsinos, S. Chatzinotas and B. Ottersten Herein, the problem of CE MIMO-OFDM precoding for transmission over frequency selective channels is tackled.
1795 Passive Intelligent Surface Assisted MIMO Powered Sustainable IoT D. Mishra and E. G. Larsson In this paper, we focus on maximizing the sum received power among the energy harvesting IoT users by jointly optimizing the active precoder for multi-antenna power beacon and the passive constant-envelope precoding based phase shifters (PS) design for PIS.
1796 Multiuser Massive Mimo Downlink Precoding Using Second-Order Spatial Sigma-Delta Modulation M. Shao, W. Ma and L. Swindlehurst In this work we study the potential of spatial S? modulation in the two-bit DAC case and under second-order modulation.
1797 Allocation of Computing Tasks In Distributed MEC Servers Co-Powered By Renewable Sources And The Power Grid D. Cecchinato, M. Berno, F. Esposito and M. Rossi We consider a Multiaccess Edge Computing (MEC) network where distributed servers have energy harvesting (e.g., solar) and storage (e.g., batteries) capabilities.
1798 Multi-Agent Deep Reinforcement Learning For Distributed Handover Management In Dense MmWave Networks M. Sana, A. De Domenico, E. C. Strinati and A. Clemente To tackle this problem, our paper proposes a deep multi-agent reinforcement learning framework for distributed handover management called RHando (Reinforced Handover).
1799 Interpretable Machine Learning In Sustainable Edge Computing: A Case Study of Short-Term Photovoltaic Power Output Prediction X. Chang, W. Li, J. Ma, T. Yang and A. Y. Zomaya In this paper, we propose a unified clustering-based prediction framework with two tree-based algorithms to provide short-term prediction of PV power output.
1800 Load Management with Predictions of Solar Energy Production for Cloud Data Centers M. Floridia, D. Lagan?, C. Mastroianni, M. Meo and D. Renga In this paper, we consider the case of Data Centers (DCs) composed of a few sites located in different geographical positions and powered with solar energy.
1801 Spectrum Allocation in Wireless Networks for Crowd Labelling X. Li, G. Zhu, K. Shen, Y. Gong and K. Huang To tackle this challenge, a novel framework of wireless crowd labelling is proposed that downloads data to many imperfect mobile annotators for repetition labelling by exploiting multicasting in wireless networks.
1802 Modeling the Environment in Deep Reinforcement Learning: The Case of Energy Harvesting Base Stations N. Piovesan, M. Miozzo and P. Dini In this paper, we focus on the design of energy self-sustainable mobile networks by enabling intelligent energy management that allows the base stations to mostly operate off-grid by using renewable energy.
1803 A Differential Approach for Rain Field Tomographic Reconstruction Using Microwave Signals from Leo Satellites X. Shen, D. D. Huang, C. Vincent, W. Wang and R. Togneri A differential approach is proposed for tomographic rain field reconstruction using the estimated signal-to-noise ratio of microwave signals from low earth orbit satellites at the ground receivers, with the unknown baseline values eliminated before using least squares to reconstruct the attenuation field.
1804 Uncertainties in Short Commercial Microwave Links Fading Due to Rain H. V. Habi and H. Messer In this paper we empirically show that while the power-law relation provides good approximation for relating attenuation and rain-rate in terrestrial microwave links of length 1-20Km, for links shorter than 1km, widely used in 5G technologies, it shows significant errors.
1805 On the Opportunistic use of Commercial Ku and Ka Band Satcom Networks for Rain Rate Estimation: Potentials and Critical Issues F. Giannetti, M. Moretti, R. Reggiannini, A. Vaccaro, S. Scarfone and A. Ortolani In this paper we study the opportunistic use of microwave satellite signals in the Ku and Ka bands to detect the presence of rain events and to estimate their intensity.
1806 Performance Analysis for Path Attenuation Estimation of Microwave Signals Due to Rainfall and Beyond B. Song, D. D. Huang, X. Shen and R. Togneri In this work, through the mean square error (MSE) analysis of an ideal RSL estimator, it is found that the estimation error can be lower than 0.01 dB for high signal-to-noise ratio (SNR), thereby making it feasible to measure other meteorological variables such as water vapor and clouds.
1807 Deep Rainrate Estimation from Highly Attenuated Downlink Signals of Ground-Based Communications Satellite Terminals K. V. Mishra, B. S. M. R. and B. Ottersten We address this problem by designing a deep convolutional neural network (CNN) that learns the relationship between the signal attenuation and rainfall rate observed by weather radars and rain gauges at a given location.
1808 Statistical Signal Processing Approach for Rain Estimation Based on Measurements from Network Management Systems J. Ostrometzky and H. Messer In this paper we apply statistical signal processing methodologies on a real-world application of using Commercial Microwave Links (CMLs) as opportunistic sensors for rain monitoring.
1809 Dynamic Oversampling in 1-Bit Quantized Asynchronous Large-Scale Multiple-Antenna Systems for Sustainable Iot Networks Z. Shao, L. T. N. Landau and R. C. de Lamare In this paper, we propose a dynamic oversampling technique for asynchronous large-scale multiple-antenna systems with 1-bit analog-to-digital converters at the base station that is suitable for sustainable internet of things and cellular networks.
1810 Dynamic Resource Allocation for Wireless Edge Machine Learning with Latency And Accuracy Guarantees M. Merluzzi, P. D. Lorenzo and S. Barbarossa In this paper, we address the problem of dynamic allocation of communication and computation resources for Edge Machine Learning (EML) exploiting Multi-Access Edge Computing (MEC).
1811 Federating Solar, Storage and Communications in the Electric Grid and Internet of things R. Ramakrishna, N. Karakoc, K. Hreinsson and A. Scaglione A futuristic infrastructure model is envisioned with distributed modules that can produce solar energy, have a storage system and provide services of lighting, electric-vehicle charging and communications.
1812 Non-Gaussian BLE-Based Indoor Localization Via Gaussian Sum Filtering Coupled with Wasserstein Distance P. Malekzadeh, S. Mehryar, P. Spachos, K. N. Plataniotis and A. Mohammadi In contrary to existing solutions, where RSSIs are assumed to have normal statistical properties, in this paper, a Gaussian Sum Filter (GSF) approach is designed to more realistically model the non-Gaussian nature of RSSIs.
1813 Optimal Joint Channel Estimation and Data Detection by L1-norm PCA for Streetscape IoT G. Sklivanitis, K. Tountas, N. Tsagkarakis, D. A. Pados and S. N. Batalama We prove, for the first time in the literature of communication theory and machine learning, the equivalence of joint maximum-likelihood (ML) optimal channel estimation and data detection (JOCEDD) to the problem of finding the L 1 -norm principal components of a real-valued data matrix.
1814 On Measuring Doppler Shifts between Tags in a Backscattering Tag-to-Tag Network with Applications in Tracking A. Ahmad et al. In this paper, we present a technique whereby passive tags can track each other in a backscattering tag-to-tag network (BTTN).
1815 Efficient Belief Propagation for Graph Matching E. Onaran and S. Villar In this short note we derive a novel belief propagation algorithm for graph matching and we numerically evaluate it in the context of matching random graphs.
1816 Supervised Graph Representation Learning for Modeling the Relationship between Structural and Functional Brain Connectivity Y. Li, R. Shafipour, G. Mateos and Z. Zhang In this paper, we propose a supervised graph representation learning method to model the relationship between brain functional connectivity (FC) and structural connectivity (SC) through a graph encoderdecoder system.
1817 Stability of Graph Neural Networks to Relative Perturbations F. Gama, A. Ribeiro and J. Bruna In this paper, we are set to study the effect that a change in the underlying graph topology that supports the signal has on the output of a GNN.
1818 Active Semi-Supervised Learning for Diffusions on Graphs B. Das, E. Isufi and G. Leus In this paper, we put forth a new paradigm for one-shot active semi-supervised learning for graph diffusions.
1819 Stochastic Graph Neural Networks Z. Gao, E. Isufi and A. Ribeiro Since stochasticity brings in a new paradigm, we develop a novel learning process for the SGNN and introduce the stochastic gradient descent (SGD) algorithm to estimate the parameters.
1820 Generative Adversarial Networks for Graph Data Imputation from Signed Observations A. Madapu, S. Segarra, S. P. Chepuri and A. G. Marques Our goal is to estimate the true underlying graph signals from our observations.
1821 Joint Frequency Domain Channel Estimation and Equalization Based on Expectation Propagation for Single Carrier Transmissions S. Sahin, A. M. Cipriano, C. Poulliat and M. Boucheret In this paper, a novel category of expectation propagation (EP) based frequency domain (FD) semi-blind receivers are proposed for single-carrier block transmissions.
1822 BP-VB-EP Based Static and Dynamic Sparse Bayesian Learning with Kronecker Structured Dictionaries C. K. Thomas and D. Slock In this paper, we propose techniques which allow handling the extension of sparse Bayesian learning (SBL) to time-varying states.
1823 Robustness of Sparse Bayesian Learning in Correlated Environments R. R. Pote and B. D. Rao In this paper we analyse the performance of Sparse Bayesian Learning (SBL) in an environment with correlated sources.
1824 A Simple Derivation of AMP and its State Evolution via First-Order Cancellation P. Schniter In this work, we provide a heuristic derivation of AMP and its state evolution, based on the idea of “firstorder cancellation,” that provides insights missing from the LBP derivation while being much shorter than the rigorous SE proof.
1825 VAMP with Vector-Valued Diagonalization R. F. H. Fischer, C. Sippel and N. Goertz In this contribution, improved versions for the update equation (“Onsager correction”) are derived from basic estimation principles.
1826 Distributed Verification of Belief Precisions Convergence in Gaussian Belief Propagation B. Li, N. Wu and Y. Wu This paper derives a simple convergence condition for Gaussian BP precisions, which can be verified in a distributed way.
1827 ADMM-Based One-Bit Quantized Signal Detection for Massive MIMO Systems With Hardware Impairments ?. T. Demir and E. Bj?rnson To solve the non-convex quadratically-constrained quadratic programming (QCQP) problem, we propose an ADMM-based algorithm with closed-form update equations.
1828 Learning Task-Based Analog-to-Digital Conversion for MIMO Receivers N. Shlezinger, R. J. G. van Sloun, I. A. M. Huijben, G. Tsintsadze and Y. C. Eldar In this work we design task-oriented analog-to-digital converters (ADCs) which operate in a data-driven manner, namely they learn how to map an analog signal into a sampled digital representation such that the system task can be efficiently carried out.
1829 One-Bit Normalized Scatter Matrix Estimation For Complex Elliptically Symmetric Distributions C. Liu and P. P. Vaidyanathan This paper shows that the arcsine law remains applicable to CES distributions.
1830 One-Bit DoA Estimation via Sparse Linear Arrays S. Sedighi, B. Shankar, M. Soltanalian and B. Ottersten In this paper, the problem of DoA estimation from one-bit measurements received by an SLA is considered and a novel framework for solving this problem is proposed.
1831 One-Bit Sampling in Fractional Fourier Domain A. Bhandari, O. Graf, F. Krahmer and A. I. Zayed In this paper, we discuss a different approach to sampling theory in the FrFT domain.
1832 Target Parameter Estimation via One-Bit PMCW Radar H. Zhu, X. Shang and J. Li We formulate the target parameter estimation problem as a sparse signal recovery problem and use the alternating direction method of multipliers (ADMM) to solve it efficiently.
1833 Mobility-Aware Beam Steering in Metasurface-Based Programmable Wireless Environments C. Liaskos, S. Nie, A. Tsioliaridou, A. Pitsillides, S. Ioannidis and I. F. Akyildiz In this work we study the effects of user device mobility on the efficiency of PWEs.
1834 Dynamic Metasurface Antennas for Bit-Constrained MIMO-OFDM Receivers H. Wang et al. In this work we study the application of DMAs for MIMO-OFDM receivers operating with bit-constrained analog-to-digital converters (ADCs).
1835 Using Intelligent Reflecting Surfaces for Rank Improvement in MIMO Communications ?. ?zdogan, E. Bj?rnson and E. G. Larsson In this paper, we study another use case that might have a larger impact on the channel capacity: rank improvement.
1836 Optimizing Backscattering Coefficient Design for Minimizing BER at Monostatic MIMO reader D. Mishra and J. Yuan We present a novel monostatic backscatter communication (BSC) protocol for multiple-input-multiple-output (MIMO) reader to detect signals from a single-antenna tag.
1837 Real-Time Implementation Aspects of Large Intelligent Surfaces H. Tataria, F. Tufvesson and O. Edfors Using the fully active LIS as the basis of our exposition, we first present a rigorous discussion on the relative merits and disadvantages of possible implementation architectures from a radio-frequency circuits and real-time processing viewpoints. We then show that a distributed architecture based on a common module interfacing a smaller number of antennas can be scalable.
1838 A Hardware Architecture For Reconfigurable Intelligent Surfaces with Minimal Active Elements for Explicit Channel Estimation G. C. Alexandropoulos and E. Vlachos In this paper, we present a novel RIS architecture comprising of any number of passive reflecting elements, a simple controller for their adjustable configuration, and a single Radio Frequency (RF) chain for baseband measurements.
1839 Conditional Density Driven Grid Design in Point-Mass Filter J. Dun?k, O. Straka and J. Matou?ek In the paper, a novel conditional density driven grid (CDDG) design is proposed.
1840 Enhanced Safety of Autonomous Driving by Incorporating Terrestrial Signals of Opportunity M. Maaref, J. Khalife and Z. M. Kassas This framework aims to incorporate terrestrial signals of opportunity (SOPs) alongside GPS signals to provide tight horizontal protection level (HPL) bounds to enhance the safety of autonomous driving.
1841 Opportunistic use of GNSS Signals to Characterize the Environment by Means of Machine Learning Based Processing F. Dovis, R. Imam, W. Qin, C. Savas and H. Visser Given this framework the present paper will describe the use of GNSS as signals of opportunity for the characterization of the operative environment by processing the GNSS observables through Machine Learning (ML) algorithms that can be used as efficient features extractors.
1842 An Optimal Symmetric Threshold Strategy for Remote Estimation Over The Collision Channel X. Zhang, M. M. Vasconcelos, W. Cui and U. Mitra An Optimal Symmetric Threshold Strategy for Remote Estimation Over The Collision Channel
1843 Automotive Collision Risk Estimation Under Cooperative Sensing D. LaChapelle, T. Humphreys, L. Narula, P. Iannucci and E. Moradi-Pari This paper offers a technique for estimating collision risk for automated ground vehicles engaged in cooperative sensing.
1844 Exploitation of 3D City Maps for Hybrid 5G RTT and GNSS Positioning Simulations J. A. del Peral-Rosado et al. To circumvent this limitation, the proposed simulation method is based on using three-dimensional (3D) city maps to coherently determine the line-of-sight (LoS) conditions of the available satellite and cellular links.
1845 A Dataset for Measuring Reading Levels In India At Scale D. Agarwal, J. Gupchup and N. Baghel To solve this assessment problem and advance deep learning research in regional Indian languages, we present the ASER dataset of children in the age group of 6-14.
1846 Noise-Robust Key-Phrase Detectors for Automated Classroom Feedback B. Zylich and J. Whitehill With the goal of giving teachers automated feedback about their classrooms, we investigate how to train automatic speech detectors of key phrases such as good job, thank you, please, and you?re welcome.
1847 Experiments in Creating Online Course Content for Signal Processing Education C. -. Jansson, R. Thottappillil, S. Hillman, S. M?ller, K. V. S. Hari and R. Sundaresan This paper presents the details and experiences of creating course content and presents guidelines for prospective content creators.
1848 Teaching Signals and Systems – A First Course in Signal Processing N. P. Rakhashi, A. A. Bhurane and V. M. Gadre Giving due consideration to this matter, this paper reflects on the experiences in teaching this course.
1849 Cochlear Signal Processing: A Platform for Learning the Fundamentals of Digital Signal Processing E. Ambikairajah and V. Sethu In this paper we propose that cochlear signal processing is an excellent platform that brings together almost all the fundamental DSP concepts taught in an introductory course and also lends itself as a suitable candidate for project-based learning.
1850 Multimodal Learning for Classroom Activity Detection H. Li et al. In this paper, we address the above challenges by using a novel attention based neural framework.
1851 Automatic Fluency Evaluation of Spontaneous Speech Using Disfluency-Based Features H. Deng, Y. Lin, T. Utsuro, A. Kobayashi, H. Nishizaki and J. Hoshino This paper describes an automatic fluency evaluation of spontaneous speech.
1852 Intelligent Student Behavior Analysis System for Real Classrooms R. Zheng, F. Jiang and R. Shen In this paper, we design an intelligent student behavior analysis system for recorded classrooms, which automatically detects hand-raising, standing, and sleeping behaviors of students.
1853 Coded Illumination and Multiplexing for Lensless Imaging Y. Zheng, R. Zhang and M. S. Asif In this paper, we propose a new method to address the problem of illconditioning by combining coded illumination patterns with the mask-based lensless imaging.
1854 Sparse Convolutional Beamforming for Wireless Ultrasound A. Mamistvalov and Y. C. Eldar In this paper we combine recently proposed methods for reducing sampling rate in each channel, and spatial reduction in the number of channels using sparse arrays, in order to implement beamforming at a low data rate without impacting image quality.
1855 Divergence-Based Adaptive Extreme Video Completion M. El Helou, R. Zhou, F. Schmutz, F. Guibert and S. S?sstrunk We propose an extension of a state-of-the-art extreme image completion algorithm to extreme video completion.
1856 Encoding and Decoding Mixed Bandlimited Signals Using Spiking Integrate-and-Fire Neurons K. Adam, A. Scholefield and M. Vetterli In this paper, we investigate the encoding and decoding potential of such a layer.
1857 On the Effect of Reflectance on Phasor Field Non-Line-of-Sight Imaging I. Guill?n, X. Liu, A. Velten, D. Gutierrez and A. Jarabo In this work we complement recent theoretical analysis of phasor field-based reconstruction, by empirically analyzing the effect of reflectance of the hidden scenes on reconstruction.
1858 Signal Sensing and Reconstruction Paradigms for a Novel Multi-Source Static Computed Tomography System A. Kowtal et al. This paper discusses new sensing and reconstruction paradigms enabled by this new CT architecture.
1859 Sampling Classes of Non-Bandlimited Signals Using Integrate-and-Fire Devices: Average Case Analysis R. Alexandru, N. T. Thao, D. Rzepka and P. L. Dragotti We investigate the use of integrate-and-fire systems to efficiently sample classes of non-bandlimited signals such as bursts of spikes.
1860 Towards an Intelligent Microscope: Adaptively Learned Illumination for Optimal Sample Classification A. Chaware, C. L. Cooke, K. Kim and R. Horstmeyer We present a reinforcement learning system that adaptively explores optimal patterns to illuminate specimens for immediate classification.
1861 High Dynamic Range Imaging Using Deep Image Priors G. Jagatap and C. Hegde In this paper, we introduce a new approach for HDR image reconstruction using neural priors that require no training data.
1862 Kernel Computations from Large-Scale Random Features Obtained by Optical Processing Units R. Ohana et al. In this paper, we show that this operation results in a dot-product kernel that has connections to the polynomial kernel, and we extend this computation to arbitrary powers of the feature map.
1863 Multi-Depth Computational Periscopy with an Ordinary Camera C. Saunders, R. Bose, J. Murray-Bruce and V. K. Goyal We demonstrate non-line-of-sight imaging of multi-depth scenes using only a single photograph from an ordinary digital camera.
1864 Super-Resolution with Noisy Measurements: Reconciling Upper and Lower Bounds H. Qiao, S. Shahsavari and P. Pal The main contribution of this paper is to derive the Cram?r-Rao Bound for noisy super- resolution problem and understand its scaling as a function of the super-resolution factor.