paper_id
stringlengths
15
35
title
stringlengths
26
182
authors
listlengths
1
25
isca_url
stringlengths
66
86
pdf_url
stringlengths
65
85
doi
stringlengths
27
30
pages
stringlengths
3
9
bibtex
large_stringlengths
294
850
abstract
large_stringlengths
247
1.59k
arxiv_id
stringlengths
10
10
arxiv_id_source
stringclasses
2 values
ali21_interspeech
Group Delay Based Re-Weighted Sparse Recovery Algorithms for Robust and High-Resolution Source Separation in DOA Framework
[ "Murtiza Ali", "Ashwani Koul", "Karan Nathwani" ]
https://www.isca-archive.org/interspeech_2021/ali21_interspeech.html
https://www.isca-archive.org/interspeech_2021/ali21_interspeech.pdf
10.21437/Interspeech.2021-164
3031-3035
@inproceedings{ali21_interspeech, title = {{Group Delay Based Re-Weighted Sparse Recovery Algorithms for Robust and High-Resolution Source Separation in DOA Framework}}, author = {Murtiza Ali and Ashwani Koul and Karan Nathwani}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3031--...
Sparse Recovery (SR) algorithms have been used widely for direction-of-arrival (DOA) estimation in spatially contiguous plane wave for their robust performance. But these algorithms have proven to be computationally costly. With a few sensors and at low SNRs, the noise dominates the data singular vectors and the sparse...
null
null
han21d_interspeech
Continuous Speech Separation Using Speaker Inventory for Long Recording
[ "Cong Han", "Yi Luo", "Chenda Li", "Tianyan Zhou", "Keisuke Kinoshita", "Shinji Watanabe", "Marc Delcroix", "Hakan Erdogan", "John R. Hershey", "Nima Mesgarani", "Zhuo Chen" ]
https://www.isca-archive.org/interspeech_2021/han21d_interspeech.html
https://www.isca-archive.org/interspeech_2021/han21d_interspeech.pdf
10.21437/Interspeech.2021-338
3036-3040
@inproceedings{han21d_interspeech, title = {{Continuous Speech Separation Using Speaker Inventory for Long Recording}}, author = {Cong Han and Yi Luo and Chenda Li and Tianyan Zhou and Keisuke Kinoshita and Shinji Watanabe and Marc Delcroix and Hakan Erdogan and John R. Hershey and Nima Mesgarani and Zhuo Ch...
Leveraging additional speaker information to facilitate speech separation has received increasing attention in recent years. Recent research includes extracting target speech by using the target speaker’s voice snippet and jointly separating all participating speakers by using a pool of additional speaker signals, whic...
2012.09727
title_judge
yuan21_interspeech
Crossfire Conditional Generative Adversarial Networks for Singing Voice Extraction
[ "Weitao Yuan", "Shengbei Wang", "Xiangrui Li", "Masashi Unoki", "Wenwu Wang" ]
https://www.isca-archive.org/interspeech_2021/yuan21_interspeech.html
https://www.isca-archive.org/interspeech_2021/yuan21_interspeech.pdf
10.21437/Interspeech.2021-433
3041-3045
@inproceedings{yuan21_interspeech, title = {{Crossfire Conditional Generative Adversarial Networks for Singing Voice Extraction}}, author = {Weitao Yuan and Shengbei Wang and Xiangrui Li and Masashi Unoki and Wenwu Wang}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3041--3045}, ...
Generative adversarial networks (GANs) and Conditional GANs (cGANs) have recently been applied for singing voice extraction (SVE), since they can accurately model the vocal distributions and effectively utilize a large amount of unlabelled datasets. However, current GANs/cGANs based SVE frameworks have no explicit mech...
null
null
wang21w_interspeech
End-to-End Speech Separation Using Orthogonal Representation in Complex and Real Time-Frequency Domain
[ "Kai Wang", "Hao Huang", "Ying Hu", "Zhihua Huang", "Sheng Li" ]
https://www.isca-archive.org/interspeech_2021/wang21w_interspeech.html
https://www.isca-archive.org/interspeech_2021/wang21w_interspeech.pdf
10.21437/Interspeech.2021-504
3046-3050
@inproceedings{wang21w_interspeech, title = {{End-to-End Speech Separation Using Orthogonal Representation in Complex and Real Time-Frequency Domain}}, author = {Kai Wang and Hao Huang and Ying Hu and Zhihua Huang and Sheng Li}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3046--3...
Traditional single channel speech separation in the time-frequency (T-F) domain often faces the problem of phase reconstruction. Due to the fact that the real-valued network is not suitable for dealing with complex-valued representation, the performance of the T-F domain speech separation method is often constrained fr...
null
null
nakagome21_interspeech
Efficient and Stable Adversarial Learning Using Unpaired Data for Unsupervised Multichannel Speech Separation
[ "Yu Nakagome", "Masahito Togami", "Tetsuji Ogawa", "Tetsunori Kobayashi" ]
https://www.isca-archive.org/interspeech_2021/nakagome21_interspeech.html
https://www.isca-archive.org/interspeech_2021/nakagome21_interspeech.pdf
10.21437/Interspeech.2021-523
3051-3055
@inproceedings{nakagome21_interspeech, title = {{Efficient and Stable Adversarial Learning Using Unpaired Data for Unsupervised Multichannel Speech Separation}}, author = {Yu Nakagome and Masahito Togami and Tetsuji Ogawa and Tetsunori Kobayashi}, year = {2021}, booktitle = {{Interspeech 2021}}, p...
This study presents a framework to enable efficient and stable adversarial learning of unsupervised multichannel source separation models. When the paired data, i.e., the mixture and the corresponding clean speech, are not available for training, it is promising to exploit generative adversarial networks (GANs), where ...
null
null
huang21h_interspeech
Stabilizing Label Assignment for Speech Separation by Self-Supervised Pre-Training
[ "Sung-Feng Huang", "Shun-Po Chuang", "Da-Rong Liu", "Yi-Chen Chen", "Gene-Ping Yang", "Hung-yi Lee" ]
https://www.isca-archive.org/interspeech_2021/huang21h_interspeech.html
https://www.isca-archive.org/interspeech_2021/huang21h_interspeech.pdf
10.21437/Interspeech.2021-763
3056-3060
@inproceedings{huang21h_interspeech, title = {{Stabilizing Label Assignment for Speech Separation by Self-Supervised Pre-Training}}, author = {Sung-Feng Huang and Shun-Po Chuang and Da-Rong Liu and Yi-Chen Chen and Gene-Ping Yang and Hung-yi Lee}, year = {2021}, booktitle = {{Interspeech 2021}}, p...
Speech separation has been well developed, with the very successful permutation invariant training (PIT) approach, although the frequent label assignment switching happening during PIT training remains to be a problem when better convergence speed and achievable performance are desired. In this paper, we propose to per...
2010.15366
title_snapshot
wang21x_interspeech
Dual-Path Filter Network: Speaker-Aware Modeling for Speech Separation
[ "Fan-Lin Wang", "Yu-Huai Peng", "Hung-Shin Lee", "Hsin-Min Wang" ]
https://www.isca-archive.org/interspeech_2021/wang21x_interspeech.html
https://www.isca-archive.org/interspeech_2021/wang21x_interspeech.pdf
10.21437/Interspeech.2021-858
3061-3065
@inproceedings{wang21x_interspeech, title = {{Dual-Path Filter Network: Speaker-Aware Modeling for Speech Separation}}, author = {Fan-Lin Wang and Yu-Huai Peng and Hung-Shin Lee and Hsin-Min Wang}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3061--3065}, doi = {10.21437/I...
Speech separation has been extensively studied to deal with the cocktail party problem in recent years. All related approaches can be divided into two categories: time-frequency domain methods and time domain methods. In addition, some methods try to generate speaker vectors to support source separation. In this study,...
2106.07579
title_snapshot
wu21f_interspeech
Investigation of Practical Aspects of Single Channel Speech Separation for ASR
[ "Jian Wu", "Zhuo Chen", "Sanyuan Chen", "Yu Wu", "Takuya Yoshioka", "Naoyuki Kanda", "Shujie Liu", "Jinyu Li" ]
https://www.isca-archive.org/interspeech_2021/wu21f_interspeech.html
https://www.isca-archive.org/interspeech_2021/wu21f_interspeech.pdf
10.21437/Interspeech.2021-921
3066-3070
@inproceedings{wu21f_interspeech, title = {{Investigation of Practical Aspects of Single Channel Speech Separation for ASR}}, author = {Jian Wu and Zhuo Chen and Sanyuan Chen and Yu Wu and Takuya Yoshioka and Naoyuki Kanda and Shujie Liu and Jinyu Li}, year = {2021}, booktitle = {{Interspeech 2021}}...
Speech separation has been successfully applied as a front-end processing module of conversation transcription systems thanks to its ability to handle overlapped speech and its flexibility to combine with downstream tasks such as automatic speech recognition (ASR). However, a speech separation model often introduces ta...
2107.01922
title_snapshot
luo21c_interspeech
Implicit Filter-and-Sum Network for End-to-End Multi-Channel Speech Separation
[ "Yi Luo", "Nima Mesgarani" ]
https://www.isca-archive.org/interspeech_2021/luo21c_interspeech.html
https://www.isca-archive.org/interspeech_2021/luo21c_interspeech.pdf
10.21437/Interspeech.2021-1158
3071-3075
@inproceedings{luo21c_interspeech, title = {{Implicit Filter-and-Sum Network for End-to-End Multi-Channel Speech Separation}}, author = {Yi Luo and Nima Mesgarani}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3071--3075}, doi = {10.21437/Interspeech.2021-1158}, issn ...
Various neural network architectures have been proposed in recent years for the task of multi-channel speech separation. Among them, the filter-and-sum network (FaSNet) performs end-to-end time-domain filter-and-sum beamforming and has shown effective in both ad-hoc and fixed microphone array geometries. However, wheth...
2011.08401
title_judge
xu21i_interspeech
Generalized Spatio-Temporal RNN Beamformer for Target Speech Separation
[ "Yong Xu", "Zhuohuang Zhang", "Meng Yu", "Shi-Xiong Zhang", "Dong Yu" ]
https://www.isca-archive.org/interspeech_2021/xu21i_interspeech.html
https://www.isca-archive.org/interspeech_2021/xu21i_interspeech.pdf
10.21437/Interspeech.2021-430
3076-3080
@inproceedings{xu21i_interspeech, title = {{Generalized Spatio-Temporal RNN Beamformer for Target Speech Separation}}, author = {Yong Xu and Zhuohuang Zhang and Meng Yu and Shi-Xiong Zhang and Dong Yu}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3076--3080}, doi = {10.21...
Although the conventional mask-based minimum variance distortionless response (MVDR) could reduce the non-linear distortion, the residual noise level of the MVDR separated speech is still high. In this paper, we propose a spatio-temporal recurrent neural network based beamformer (RNN-BF) for target speech separation. T...
2101.01280
title_snapshot
liu21j_interspeech
End-to-End Neural Diarization: From Transformer to Conformer
[ "Yi Chieh Liu", "Eunjung Han", "Chul Lee", "Andreas Stolcke" ]
https://www.isca-archive.org/interspeech_2021/liu21j_interspeech.html
https://www.isca-archive.org/interspeech_2021/liu21j_interspeech.pdf
10.21437/Interspeech.2021-1909
3081-3085
@inproceedings{liu21j_interspeech, title = {{End-to-End Neural Diarization: From Transformer to Conformer}}, author = {Yi Chieh Liu and Eunjung Han and Chul Lee and Andreas Stolcke}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3081--3085}, doi = {10.21437/Interspeech.2021...
We propose a new end-to-end neural diarization (EEND) system that is based on Conformer, a recently proposed neural architecture that combines convolutional mappings and Transformer to model both local and global dependencies in speech. We first show that data augmentation and convolutional subsampling layers enhance t...
2106.07167
title_snapshot
jung21_interspeech
Three-Class Overlapped Speech Detection Using a Convolutional Recurrent Neural Network
[ "Jee-weon Jung", "Hee-Soo Heo", "Youngki Kwon", "Joon Son Chung", "Bong-Jin Lee" ]
https://www.isca-archive.org/interspeech_2021/jung21_interspeech.html
https://www.isca-archive.org/interspeech_2021/jung21_interspeech.pdf
10.21437/Interspeech.2021-149
3086-3090
@inproceedings{jung21_interspeech, title = {{Three-Class Overlapped Speech Detection Using a Convolutional Recurrent Neural Network}}, author = {Jee-weon Jung and Hee-Soo Heo and Youngki Kwon and Joon Son Chung and Bong-Jin Lee}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3086--...
In this work, we propose an overlapped speech detection system trained as a three-class classifier. Unlike conventional systems that perform binary classification as to whether or not a frame contains overlapped speech, the proposed approach classifies into three classes: non-speech, single speaker speech, and overlapp...
2104.02878
title_snapshot
wan21_interspeech
Online Speaker Diarization Equipped with Discriminative Modeling and Guided Inference
[ "Xucheng Wan", "Kai Liu", "Huan Zhou" ]
https://www.isca-archive.org/interspeech_2021/wan21_interspeech.html
https://www.isca-archive.org/interspeech_2021/wan21_interspeech.pdf
10.21437/Interspeech.2021-261
3091-3095
@inproceedings{wan21_interspeech, title = {{Online Speaker Diarization Equipped with Discriminative Modeling and Guided Inference}}, author = {Xucheng Wan and Kai Liu and Huan Zhou}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3091--3095}, doi = {10.21437/Interspeech.2021...
Despite considerable efforts, online speaker diarization remains an ongoing challenge. In this study, we propose to tackle the challenge from two perspectives, to endow diarization model with discriminability and to rectify less-reliable online inference with guidance. Specifically, based on the current prior art, UIS-...
null
null
takashima21_interspeech
Semi-Supervised Training with Pseudo-Labeling for End-To-End Neural Diarization
[ "Yuki Takashima", "Yusuke Fujita", "Shota Horiguchi", "Shinji Watanabe", "Leibny Paola García Perera", "Kenji Nagamatsu" ]
https://www.isca-archive.org/interspeech_2021/takashima21_interspeech.html
https://www.isca-archive.org/interspeech_2021/takashima21_interspeech.pdf
10.21437/Interspeech.2021-384
3096-3100
@inproceedings{takashima21_interspeech, title = {{Semi-Supervised Training with Pseudo-Labeling for End-To-End Neural Diarization}}, author = {Yuki Takashima and Yusuke Fujita and Shota Horiguchi and Shinji Watanabe and Leibny Paola García Perera and Kenji Nagamatsu}, year = {2021}, booktitle = {{In...
In this paper, we present a semi-supervised training technique using pseudo-labeling for end-to-end neural diarization (EEND). The EEND system has shown promising performance compared with traditional clustering-based methods, especially in the case of overlapping speech. However, to get a well-tuned model, EEND requir...
2106.04764
title_snapshot
kwon21b_interspeech
Adapting Speaker Embeddings for Speaker Diarisation
[ "Youngki Kwon", "Jee-weon Jung", "Hee-Soo Heo", "You Jin Kim", "Bong-Jin Lee", "Joon Son Chung" ]
https://www.isca-archive.org/interspeech_2021/kwon21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/kwon21b_interspeech.pdf
10.21437/Interspeech.2021-448
3101-3105
@inproceedings{kwon21b_interspeech, title = {{Adapting Speaker Embeddings for Speaker Diarisation}}, author = {Youngki Kwon and Jee-weon Jung and Hee-Soo Heo and You Jin Kim and Bong-Jin Lee and Joon Son Chung}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3101--3105}, doi ...
The goal of this paper is to adapt speaker embeddings for solving the problem of speaker diarisation. The quality of speaker embeddings is paramount to the performance of speaker diarisation systems. Despite this, prior works in the field have directly used embeddings designed only to be effective on the speaker verifi...
2104.02879
title_snapshot
wang21y_interspeech
Scenario-Dependent Speaker Diarization for DIHARD-III Challenge
[ "Yu-Xuan Wang", "Jun Du", "Maokui He", "Shu-Tong Niu", "Lei Sun", "Chin-Hui Lee" ]
https://www.isca-archive.org/interspeech_2021/wang21y_interspeech.html
https://www.isca-archive.org/interspeech_2021/wang21y_interspeech.pdf
10.21437/Interspeech.2021-516
3106-3110
@inproceedings{wang21y_interspeech, title = {{Scenario-Dependent Speaker Diarization for DIHARD-III Challenge}}, author = {Yu-Xuan Wang and Jun Du and Maokui He and Shu-Tong Niu and Lei Sun and Chin-Hui Lee}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3106--3110}, doi = ...
In this study, we propose a scenario-dependent speaker diarization approach to handling the diversified scenarios of 11 domains encountered in DIHARD-III challenge with a divide-and-conquer strategy. First, using a ResNet-based audio domain classifier, all domains in DIHARD-III challenge could be divided into several s...
null
null
bredin21_interspeech
End-To-End Speaker Segmentation for Overlap-Aware Resegmentation
[ "Hervé Bredin", "Antoine Laurent" ]
https://www.isca-archive.org/interspeech_2021/bredin21_interspeech.html
https://www.isca-archive.org/interspeech_2021/bredin21_interspeech.pdf
10.21437/Interspeech.2021-560
3111-3115
@inproceedings{bredin21_interspeech, title = {{End-To-End Speaker Segmentation for Overlap-Aware Resegmentation}}, author = {Hervé Bredin and Antoine Laurent}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3111--3115}, doi = {10.21437/Interspeech.2021-560}, issn = {2...
Speaker segmentation consists in partitioning a conversation between one or more speakers into speaker turns. Usually addressed as the late combination of three sub-tasks (voice activity detection, speaker change detection, and overlapped speech detection), we propose to train an end-to-end segmentation model that does...
2104.04045
title_snapshot
xue21d_interspeech
Online Streaming End-to-End Neural Diarization Handling Overlapping Speech and Flexible Numbers of Speakers
[ "Yawen Xue", "Shota Horiguchi", "Yusuke Fujita", "Yuki Takashima", "Shinji Watanabe", "Leibny Paola García Perera", "Kenji Nagamatsu" ]
https://www.isca-archive.org/interspeech_2021/xue21d_interspeech.html
https://www.isca-archive.org/interspeech_2021/xue21d_interspeech.pdf
10.21437/Interspeech.2021-708
3116-3120
@inproceedings{xue21d_interspeech, title = {{Online Streaming End-to-End Neural Diarization Handling Overlapping Speech and Flexible Numbers of Speakers}}, author = {Yawen Xue and Shota Horiguchi and Yusuke Fujita and Yuki Takashima and Shinji Watanabe and Leibny Paola García Perera and Kenji Nagamatsu}, y...
We propose a streaming diarization method based on an end-to-end neural diarization (EEND) model, which handles flexible numbers of speakers and overlapping speech. In our previous study, the speaker-tracing buffer (STB) mechanism was proposed to achieve a chunk-wise streaming diarization using a pre-trained EEND model...
2101.08473
title_snapshot
anidjar21_interspeech
A Thousand Words are Worth More Than One Recording: Based Speaker Change Detection
[ "Or Haim Anidjar", "Itshak Lapidot", "Chen Hajaj", "Amit Dvir" ]
https://www.isca-archive.org/interspeech_2021/anidjar21_interspeech.html
https://www.isca-archive.org/interspeech_2021/anidjar21_interspeech.pdf
10.21437/Interspeech.2021-87
3121-3125
@inproceedings{anidjar21_interspeech, title = {{A Thousand Words are Worth More Than One Recording: Word-Embedding Based Speaker Change Detection}}, author = {Or Haim Anidjar and Itshak Lapidot and Chen Hajaj and Amit Dvir}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3121--3125}...
Speaker Change Detection (SCD) is the task of segmenting an input audio-recording according to speaker interchanges. This task is essential for many applications, such as automatic voice transcription or Speaker Diarization (SD). This paper focuses on the essential task of audio segmentation and suggests a word-embeddi...
2006.01206
title_judge
futamata21_interspeech
Phrase Break Prediction with Bidirectional Encoder Representations in Japanese Text-to-Speech Synthesis
[ "Kosuke Futamata", "Byeongseon Park", "Ryuichi Yamamoto", "Kentaro Tachibana" ]
https://www.isca-archive.org/interspeech_2021/futamata21_interspeech.html
https://www.isca-archive.org/interspeech_2021/futamata21_interspeech.pdf
10.21437/Interspeech.2021-252
3126-3130
@inproceedings{futamata21_interspeech, title = {{Phrase Break Prediction with Bidirectional Encoder Representations in Japanese Text-to-Speech Synthesis}}, author = {Kosuke Futamata and Byeongseon Park and Ryuichi Yamamoto and Kentaro Tachibana}, year = {2021}, booktitle = {{Interspeech 2021}}, pa...
We propose a novel phrase break prediction method that combines implicit features extracted from a pre-trained large language model, a.k.a BERT, and explicit features extracted from BiLSTM with linguistic features. In conventional BiLSTM-based methods, word representations and/or sentence representations are used as in...
2104.12395
title_snapshot
vallesperez21_interspeech
Improving Multi-Speaker TTS Prosody Variance with a Residual Encoder and Normalizing Flows
[ "Iván Vallés-Pérez", "Julian Roth", "Grzegorz Beringer", "Roberto Barra-Chicote", "Jasha Droppo" ]
https://www.isca-archive.org/interspeech_2021/vallesperez21_interspeech.html
https://www.isca-archive.org/interspeech_2021/vallesperez21_interspeech.pdf
10.21437/Interspeech.2021-562
3131-3135
@inproceedings{vallesperez21_interspeech, title = {{Improving Multi-Speaker TTS Prosody Variance with a Residual Encoder and Normalizing Flows}}, author = {Iván Vallés-Pérez and Julian Roth and Grzegorz Beringer and Roberto Barra-Chicote and Jasha Droppo}, year = {2021}, booktitle = {{Interspeech 20...
Text-to-speech systems recently achieved almost indistinguishable quality from human speech. However, the prosody of those systems is generally flatter than natural speech, producing samples with low expressiveness. Disentanglement of speaker id and prosody is crucial in text-to-speech systems to improve on naturalness...
2106.05762
title_snapshot
du21b_interspeech
Rich Prosody Diversity Modelling with Phone-Level Mixture Density Network
[ "Chenpeng Du", "Kai Yu" ]
https://www.isca-archive.org/interspeech_2021/du21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/du21b_interspeech.pdf
10.21437/Interspeech.2021-802
3136-3140
@inproceedings{du21b_interspeech, title = {{Rich Prosody Diversity Modelling with Phone-Level Mixture Density Network}}, author = {Chenpeng Du and Kai Yu}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3136--3140}, doi = {10.21437/Interspeech.2021-802}, issn = {2958-...
Generating natural speech with a diverse and smooth prosody pattern is a challenging task. Although random sampling with phone-level prosody distribution has been investigated to generate different prosody patterns, the diversity of the generated speech is still very limited and far from what can be achieved by humans....
2102.00851
title_snapshot
fujita21_interspeech
Phoneme Duration Modeling Using Speech Rhythm-Based Speaker Embeddings for Multi-Speaker Speech Synthesis
[ "Kenichi Fujita", "Atsushi Ando", "Yusuke Ijima" ]
https://www.isca-archive.org/interspeech_2021/fujita21_interspeech.html
https://www.isca-archive.org/interspeech_2021/fujita21_interspeech.pdf
10.21437/Interspeech.2021-826
3141-3145
@inproceedings{fujita21_interspeech, title = {{Phoneme Duration Modeling Using Speech Rhythm-Based Speaker Embeddings for Multi-Speaker Speech Synthesis}}, author = {Kenichi Fujita and Atsushi Ando and Yusuke Ijima}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3141--3145}, doi ...
This paper proposes a novel speech-rhythm-based method for speaker embeddings. Conventionally spectral feature-based speaker embedding vectors such as the x-vector are used as auxiliary information for multi-speaker speech synthesis. However, speech synthesis with conventional embeddings has difficulty reproducing the ...
null
null
zou21_interspeech
Fine-Grained Prosody Modeling in Neural Speech Synthesis Using ToBI Representation
[ "Yuxiang Zou", "Shichao Liu", "Xiang Yin", "Haopeng Lin", "Chunfeng Wang", "Haoyu Zhang", "Zejun Ma" ]
https://www.isca-archive.org/interspeech_2021/zou21_interspeech.html
https://www.isca-archive.org/interspeech_2021/zou21_interspeech.pdf
10.21437/Interspeech.2021-883
3146-3150
@inproceedings{zou21_interspeech, title = {{Fine-Grained Prosody Modeling in Neural Speech Synthesis Using ToBI Representation}}, author = {Yuxiang Zou and Shichao Liu and Xiang Yin and Haopeng Lin and Chunfeng Wang and Haoyu Zhang and Zejun Ma}, year = {2021}, booktitle = {{Interspeech 2021}}, pa...
Benefiting from the great development of deep learning, modern neural text-to-speech (TTS) models can generate speech indistinguishable from natural speech. However, The generated utterances often keep an average prosodic style of the database instead of having rich prosodic variation. For pitch-stressed languages, suc...
null
null
sharma21b_interspeech
Intra-Sentential Speaking Rate Control in Neural Text-To-Speech for Automatic Dubbing
[ "Mayank Sharma", "Yogesh Virkar", "Marcello Federico", "Roberto Barra-Chicote", "Robert Enyedi" ]
https://www.isca-archive.org/interspeech_2021/sharma21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/sharma21b_interspeech.pdf
10.21437/Interspeech.2021-1012
3151-3155
@inproceedings{sharma21b_interspeech, title = {{Intra-Sentential Speaking Rate Control in Neural Text-To-Speech for Automatic Dubbing}}, author = {Mayank Sharma and Yogesh Virkar and Marcello Federico and Roberto Barra-Chicote and Robert Enyedi}, year = {2021}, booktitle = {{Interspeech 2021}}, pa...
Automatically dubbed speech of a video involves: (i) segmenting the target sentences into phrases to reflect the speech-pause arrangement used by the original speaker, and (ii) adjusting the speaking rate of the synthetic voice at the phrase-level to match the exact timing of each corresponding source phrase. In this w...
null
null
zhang21u_interspeech
Applying the Information Bottleneck Principle to Prosodic Representation Learning
[ "Guangyan Zhang", "Ying Qin", "Daxin Tan", "Tan Lee" ]
https://www.isca-archive.org/interspeech_2021/zhang21u_interspeech.html
https://www.isca-archive.org/interspeech_2021/zhang21u_interspeech.pdf
10.21437/Interspeech.2021-1049
3156-3160
@inproceedings{zhang21u_interspeech, title = {{Applying the Information Bottleneck Principle to Prosodic Representation Learning}}, author = {Guangyan Zhang and Ying Qin and Daxin Tan and Tan Lee}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3156--3160}, doi = {10.21437/I...
This paper describes a novel design of a neural network-based speech generation model for learning prosodic representation. The problem of representation learning is formulated according to the information bottleneck (IB) principle. A modified VQ-VAE quantized layer is incorporated in the speech generation model to con...
2108.02821
title_snapshot
baird21_interspeech
A Prototypical Network Approach for Evaluating Generated Emotional Speech
[ "Alice Baird", "Silvan Mertes", "Manuel Milling", "Lukas Stappen", "Thomas Wiest", "Elisabeth André", "Björn W. Schuller" ]
https://www.isca-archive.org/interspeech_2021/baird21_interspeech.html
https://www.isca-archive.org/interspeech_2021/baird21_interspeech.pdf
10.21437/Interspeech.2021-1123
3161-3165
@inproceedings{baird21_interspeech, title = {{A Prototypical Network Approach for Evaluating Generated Emotional Speech}}, author = {Alice Baird and Silvan Mertes and Manuel Milling and Lukas Stappen and Thomas Wiest and Elisabeth André and Björn W. Schuller}, year = {2021}, booktitle = {{Interspeec...
The collection of emotional speech data is a time-consuming and costly endeavour. Generative networks can be applied to augment the limited audio data artificially. However, it is challenging to evaluate generated audio for its similarity to source data, as current quantitative metrics are not necessarily suited to the...
null
null
yoshinaga21_interspeech
A Simplified Model for the Vocal Tract of [s] with Inclined Incisors
[ "Tsukasa Yoshinaga", "Kohei Tada", "Kazunori Nozaki", "Akiyoshi Iida" ]
https://www.isca-archive.org/interspeech_2021/yoshinaga21_interspeech.html
https://www.isca-archive.org/interspeech_2021/yoshinaga21_interspeech.pdf
10.21437/Interspeech.2021-231
3166-3170
@inproceedings{yoshinaga21_interspeech, title = {{A Simplified Model for the Vocal Tract of [s] with Inclined Incisors}}, author = {Tsukasa Yoshinaga and Kohei Tada and Kazunori Nozaki and Akiyoshi Iida}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3166--3170}, doi = {10....
To examine the effects of inclined incisors on the phonation of [s], a simplified vocal tract model is proposed, and the acoustic characteristics with different maxillary incisor angles are predicted by the model. As a control model, a realistic vocal tract replica of [s] was constructed from medical images, and the an...
null
null
arai21b_interspeech
Vocal-Tract Models to Visualize the Airstream of Human Breath and Droplets While Producing Speech
[ "Takayuki Arai" ]
https://www.isca-archive.org/interspeech_2021/arai21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/arai21b_interspeech.pdf
10.21437/Interspeech.2021-449
3171-3175
@inproceedings{arai21b_interspeech, title = {{Vocal-Tract Models to Visualize the Airstream of Human Breath and Droplets While Producing Speech}}, author = {Takayuki Arai}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3171--3175}, doi = {10.21437/Interspeech.2021-449}, i...
Due to the COVID-19 pandemic, visualizing the airstream of human breath during speech production has become extremely important from the viewpoint of preventing infection. In addition, visualizing droplets and the larger drops expelled when we speak consonantal sounds may help for the same reason. One visualization tec...
null
null
tanji21_interspeech
Using Transposed Convolution for Articulatory-to-Acoustic Conversion from Real-Time MRI Data
[ "Ryo Tanji", "Hidefumi Ohmura", "Kouichi Katsurada" ]
https://www.isca-archive.org/interspeech_2021/tanji21_interspeech.html
https://www.isca-archive.org/interspeech_2021/tanji21_interspeech.pdf
10.21437/Interspeech.2021-906
3176-3180
@inproceedings{tanji21_interspeech, title = {{Using Transposed Convolution for Articulatory-to-Acoustic Conversion from Real-Time MRI Data}}, author = {Ryo Tanji and Hidefumi Ohmura and Kouichi Katsurada}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3176--3180}, doi = {10...
We herein propose a deep neural network-based model for articulatory-to-acoustic conversion from real-time MRI data. Although rtMRI, which can record entire articulatory organs with a high resolution, has an advantage in articulatory-to-acoustic conversion, it has a relatively low sampling rate. To address this, we inc...
null
null
inaam21_interspeech
Comparison Between Lumped-Mass Modeling and Flow Simulation of the Reed-Type Artificial Vocal Fold
[ "Rafia Inaam", "Tsukasa Yoshinaga", "Takayuki Arai", "Hiroshi Yokoyama", "Akiyoshi Iida" ]
https://www.isca-archive.org/interspeech_2021/inaam21_interspeech.html
https://www.isca-archive.org/interspeech_2021/inaam21_interspeech.pdf
10.21437/Interspeech.2021-929
3181-3185
@inproceedings{inaam21_interspeech, title = {{Comparison Between Lumped-Mass Modeling and Flow Simulation of the Reed-Type Artificial Vocal Fold}}, author = {Rafia Inaam and Tsukasa Yoshinaga and Takayuki Arai and Hiroshi Yokoyama and Akiyoshi Iida}, year = {2021}, booktitle = {{Interspeech 2021}}, ...
The sound generated by a reed-type artificial vocal fold was predicted by a one-mass modeling and numerical flow simulation to examine the sound generation mechanisms of the artificial vocal fold. For the one-mass modeling, the reed oscillation was modeled with an equivalent spring constant, and the flow rate was estim...
null
null
werner21_interspeech
Inhalations in Speech: Acoustic and Physiological Characteristics
[ "Raphael Werner", "Susanne Fuchs", "Jürgen Trouvain", "Bernd Möbius" ]
https://www.isca-archive.org/interspeech_2021/werner21_interspeech.html
https://www.isca-archive.org/interspeech_2021/werner21_interspeech.pdf
10.21437/Interspeech.2021-1262
3186-3190
@inproceedings{werner21_interspeech, title = {{Inhalations in Speech: Acoustic and Physiological Characteristics}}, author = {Raphael Werner and Susanne Fuchs and Jürgen Trouvain and Bernd Möbius}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3186--3190}, doi = {10.21437/I...
This paper examines the acoustic properties of breath noises in speech pauses in relation to similar speech segments and with regard to their inhalation speed. We measured intensity, center of gravity, and formants, as well as kinematic data (via Respiratory Inductance Plethysmography) for inhalations, aspirations of s...
null
null
xu21j_interspeech
Model-Based Exploration of Linking Between Vowel Articulatory Space and Acoustic Space
[ "Anqi Xu", "Daniel van Niekerk", "Branislav Gerazov", "Paul Konstantin Krug", "Santitham Prom-on", "Peter Birkholz", "Yi Xu" ]
https://www.isca-archive.org/interspeech_2021/xu21j_interspeech.html
https://www.isca-archive.org/interspeech_2021/xu21j_interspeech.pdf
10.21437/Interspeech.2021-1422
3191-3195
@inproceedings{xu21j_interspeech, title = {{Model-Based Exploration of Linking Between Vowel Articulatory Space and Acoustic Space}}, author = {Anqi Xu and Daniel van Niekerk and Branislav Gerazov and Paul Konstantin Krug and Santitham Prom-on and Peter Birkholz and Yi Xu}, year = {2021}, booktitle ...
While the acoustic vowel space has been extensively studied in previous research, little is known about the high-dimensional articulatory space of vowels. The articulatory imaging techniques are limited to tracking only a few key articulators, leaving the rest of the articulators unmonitored. In the present study, we a...
null
null
elmers21_interspeech
Take a Breath: Respiratory Sounds Improve Recollection in Synthetic Speech
[ "Mikey Elmers", "Raphael Werner", "Beeke Muhlack", "Bernd Möbius", "Jürgen Trouvain" ]
https://www.isca-archive.org/interspeech_2021/elmers21_interspeech.html
https://www.isca-archive.org/interspeech_2021/elmers21_interspeech.pdf
10.21437/Interspeech.2021-1496
3196-3200
@inproceedings{elmers21_interspeech, title = {{Take a Breath: Respiratory Sounds Improve Recollection in Synthetic Speech}}, author = {Mikey Elmers and Raphael Werner and Beeke Muhlack and Bernd Möbius and Jürgen Trouvain}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3196--3200},...
This study revisits Whalen et al. (1995, JASA) by evaluating English speaking participants in a perception experiment to determine if their recollection is affected by including breath noises in sentences generated by a speech synthesis system. Whalen found an improvement in recollection for sentences that were precede...
null
null
chen21m_interspeech
Modeling Sensorimotor Adaptation in Speech Through Alterations to Forward and Inverse Models
[ "Taijing Chen", "Adam Lammert", "Benjamin Parrell" ]
https://www.isca-archive.org/interspeech_2021/chen21m_interspeech.html
https://www.isca-archive.org/interspeech_2021/chen21m_interspeech.pdf
10.21437/Interspeech.2021-1746
3201-3205
@inproceedings{chen21m_interspeech, title = {{Modeling Sensorimotor Adaptation in Speech Through Alterations to Forward and Inverse Models}}, author = {Taijing Chen and Adam Lammert and Benjamin Parrell}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3201--3205}, doi = {10....
When speakers are exposed to auditory feedback perturbations of a particular vowel, they not only adapt their productions of that vowel but also transfer this change to other, untrained, vowels. However, current models of speech sensorimotor adaptation, which rely on changes in the feedforward control of specific speec...
null
null
kawahara21_interspeech
Mixture of Orthogonal Sequences Made from Extended Time-Stretched Pulses Enables Measurement of Involuntary Voice Fundamental Frequency Response to Pitch Perturbation
[ "Hideki Kawahara", "Toshie Matsui", "Kohei Yatabe", "Ken-Ichi Sakakibara", "Minoru Tsuzaki", "Masanori Morise", "Toshio Irino" ]
https://www.isca-archive.org/interspeech_2021/kawahara21_interspeech.html
https://www.isca-archive.org/interspeech_2021/kawahara21_interspeech.pdf
10.21437/Interspeech.2021-2073
3206-3210
@inproceedings{kawahara21_interspeech, title = {{Mixture of Orthogonal Sequences Made from Extended Time-Stretched Pulses Enables Measurement of Involuntary Voice Fundamental Frequency Response to Pitch Perturbation}}, author = {Hideki Kawahara and Toshie Matsui and Kohei Yatabe and Ken-Ichi Sakakibara and M...
Auditory feedback plays an essential role in the regulation of the fundamental frequency of voiced sounds. The fundamental frequency also responds to auditory stimulation other than the speaker’s voice. We propose to use this response of the fundamental frequency of sustained vowels to frequency-modulated test signals ...
2104.01444
title_snapshot
you21c_interspeech
Contextualized Attention-Based Knowledge Transfer for Spoken Conversational Question Answering
[ "Chenyu You", "Nuo Chen", "Yuexian Zou" ]
https://www.isca-archive.org/interspeech_2021/you21c_interspeech.html
https://www.isca-archive.org/interspeech_2021/you21c_interspeech.pdf
10.21437/Interspeech.2021-110
3211-3215
@inproceedings{you21c_interspeech, title = {{Contextualized Attention-Based Knowledge Transfer for Spoken Conversational Question Answering}}, author = {Chenyu You and Nuo Chen and Yuexian Zou}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3211--3215}, doi = {10.21437/Inte...
Spoken conversational question answering (SCQA) requires machines to model the flow of multi-turn conversation given the speech utterances and text corpora. Different from traditional text question answering (QA) tasks, SCQA involves audio signal processing, passage comprehension, and contextual understanding. However,...
2010.11066
title_snapshot
duan21_interspeech
Injecting Descriptive Meta-Information into Pre-Trained Language Models with Hypernetworks
[ "Wenying Duan", "Xiaoxi He", "Zimu Zhou", "Hong Rao", "Lothar Thiele" ]
https://www.isca-archive.org/interspeech_2021/duan21_interspeech.html
https://www.isca-archive.org/interspeech_2021/duan21_interspeech.pdf
10.21437/Interspeech.2021-229
3216-3220
@inproceedings{duan21_interspeech, title = {{Injecting Descriptive Meta-Information into Pre-Trained Language Models with Hypernetworks}}, author = {Wenying Duan and Xiaoxi He and Zimu Zhou and Hong Rao and Lothar Thiele}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3216--3220}, ...
Pre-trained language models have been widely adopted as backbones in various natural language processing tasks. However, existing pre-trained language models ignore the descriptive meta-information in the text such as the distinction between the title and the mainbody, leading to over-weighted attention to insignifican...
null
null
rohmatillah21_interspeech
Causal Confusion Reduction for Robust Multi-Domain Dialogue Policy
[ "Mahdin Rohmatillah", "Jen-Tzung Chien" ]
https://www.isca-archive.org/interspeech_2021/rohmatillah21_interspeech.html
https://www.isca-archive.org/interspeech_2021/rohmatillah21_interspeech.pdf
10.21437/Interspeech.2021-534
3221-3225
@inproceedings{rohmatillah21_interspeech, title = {{Causal Confusion Reduction for Robust Multi-Domain Dialogue Policy}}, author = {Mahdin Rohmatillah and Jen-Tzung Chien}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3221--3225}, doi = {10.21437/Interspeech.2021-534}, i...
In the multi-domain dialogue system, dialog policy plays an important role since it determines the suitable actions based on the user’s goals. However, in many recent works, most of the dialogue optimizations, especially that use reinforcement learning (RL) methods, do not perform well. The main problem is that the ini...
null
null
fujie21_interspeech
Timing Generating Networks: Neural Network Based Precise Turn-Taking Timing Prediction in Multiparty Conversation
[ "Shinya Fujie", "Hayato Katayama", "Jin Sakuma", "Tetsunori Kobayashi" ]
https://www.isca-archive.org/interspeech_2021/fujie21_interspeech.html
https://www.isca-archive.org/interspeech_2021/fujie21_interspeech.pdf
10.21437/Interspeech.2021-874
3226-3230
@inproceedings{fujie21_interspeech, title = {{Timing Generating Networks: Neural Network Based Precise Turn-Taking Timing Prediction in Multiparty Conversation}}, author = {Shinya Fujie and Hayato Katayama and Jin Sakuma and Tetsunori Kobayashi}, year = {2021}, booktitle = {{Interspeech 2021}}, pa...
A brand new neural network based precise timing generation framework, named the Timing Generating Network (TGN), is proposed and applied to turn-taking timing decision problems. Although turn-taking problems have conventionally been formalized as users’ end-of-turn detection, this approach cannot estimate the precise t...
null
null
chen21n_interspeech
Human-to-Human Conversation Dataset for Learning Fine-Grained Turn-Taking Action
[ "Kehan Chen", "Zezhong Li", "Suyang Dai", "Wei Zhou", "Haiqing Chen" ]
https://www.isca-archive.org/interspeech_2021/chen21n_interspeech.html
https://www.isca-archive.org/interspeech_2021/chen21n_interspeech.pdf
10.21437/Interspeech.2021-994
3231-3235
@inproceedings{chen21n_interspeech, title = {{Human-to-Human Conversation Dataset for Learning Fine-Grained Turn-Taking Action}}, author = {Kehan Chen and Zezhong Li and Suyang Dai and Wei Zhou and Haiqing Chen}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3231--3235}, doi ...
Conducting natural turn-taking behavior takes a crucial part in the user experience of modern spoken dialogue systems. One way to build such system is to learn those behaviors from real-world human-to-human dialogues, which have the most diverse and fine-grained turn-taking actions than any manual constructed sessions....
null
null
sundararaman21_interspeech
PhonemeBERT: Joint Language Modelling of Phoneme Sequence and ASR Transcript
[ "Mukuntha Narayanan Sundararaman", "Ayush Kumar", "Jithendra Vepa" ]
https://www.isca-archive.org/interspeech_2021/sundararaman21_interspeech.html
https://www.isca-archive.org/interspeech_2021/sundararaman21_interspeech.pdf
10.21437/Interspeech.2021-1582
3236-3240
@inproceedings{sundararaman21_interspeech, title = {{PhonemeBERT: Joint Language Modelling of Phoneme Sequence and ASR Transcript}}, author = {Mukuntha Narayanan Sundararaman and Ayush Kumar and Jithendra Vepa}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3236--3240}, doi ...
Recent years have witnessed significant improvement in ASR systems to recognize spoken utterances. However, it is still a challenging task for noisy and out-of-domain data, where ASR errors are prevalent in the transcribed text. These errors significantly degrade the performance of downstream tasks such as intent and s...
2102.00804
title_judge
luo21d_interspeech
Joint Retrieval-Extraction Training for Evidence-Aware Dialog Response Selection
[ "Hongyin Luo", "James Glass", "Garima Lalwani", "Yi Zhang", "Shang-Wen Li" ]
https://www.isca-archive.org/interspeech_2021/luo21d_interspeech.html
https://www.isca-archive.org/interspeech_2021/luo21d_interspeech.pdf
10.21437/Interspeech.2021-1689
3241-3245
@inproceedings{luo21d_interspeech, title = {{Joint Retrieval-Extraction Training for Evidence-Aware Dialog Response Selection}}, author = {Hongyin Luo and James Glass and Garima Lalwani and Yi Zhang and Shang-Wen Li}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3241--3245}, doi...
Neural dialog response selection models infer by scoring each candidate response given the dialog context, and the cross-encoder method yields state-of-the-art (SOTA) results for the task. In the method, the candidate scores are computed by feeding the output embedding of the first token in the input sequence, which is...
null
null
shenoy21_interspeech
Adapting Long Context NLM for ASR Rescoring in Conversational Agents
[ "Ashish Shenoy", "Sravan Bodapati", "Monica Sunkara", "Srikanth Ronanki", "Katrin Kirchhoff" ]
https://www.isca-archive.org/interspeech_2021/shenoy21_interspeech.html
https://www.isca-archive.org/interspeech_2021/shenoy21_interspeech.pdf
10.21437/Interspeech.2021-1849
3246-3250
@inproceedings{shenoy21_interspeech, title = {{Adapting Long Context NLM for ASR Rescoring in Conversational Agents}}, author = {Ashish Shenoy and Sravan Bodapati and Monica Sunkara and Srikanth Ronanki and Katrin Kirchhoff}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3246--3250...
Neural Language Models (NLM), when trained and evaluated with context spanning multiple utterances, have been shown to consistently outperform both conventional n-gram language models and NLMs that use limited context. In this paper, we investigate various techniques to incorporate turn based context history into both ...
2104.11070
title_snapshot
li21h_interspeech
Oriental Language Recognition (OLR) 2020: Summary and Analysis
[ "Jing Li", "Binling Wang", "Yiming Zhi", "Zheng Li", "Lin Li", "Qingyang Hong", "Dong Wang" ]
https://www.isca-archive.org/interspeech_2021/li21h_interspeech.html
https://www.isca-archive.org/interspeech_2021/li21h_interspeech.pdf
10.21437/Interspeech.2021-2171
3251-3255
@inproceedings{li21h_interspeech, title = {{Oriental Language Recognition (OLR) 2020: Summary and Analysis}}, author = {Jing Li and Binling Wang and Yiming Zhi and Zheng Li and Lin Li and Qingyang Hong and Dong Wang}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3251--3255}, doi...
The fifth Oriental Language Recognition (OLR) Challenge focuses on language recognition in a variety of complex environments to promote its development. The OLR 2020 Challenge includes three tasks: (1) cross-channel language identification, (2) dialect identification, and (3) noisy language identification. We choose C ...
2107.05365
title_snapshot
duroselle21b_interspeech
Language Recognition on Unknown Conditions: The LORIA-Inria-MULTISPEECH System for AP20-OLR Challenge
[ "Raphaël Duroselle", "Md. Sahidullah", "Denis Jouvet", "Irina Illina" ]
https://www.isca-archive.org/interspeech_2021/duroselle21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/duroselle21b_interspeech.pdf
10.21437/Interspeech.2021-276
3256-3260
@inproceedings{duroselle21b_interspeech, title = {{Language Recognition on Unknown Conditions: The LORIA-Inria-MULTISPEECH System for AP20-OLR Challenge}}, author = {Raphaël Duroselle and Md. Sahidullah and Denis Jouvet and Irina Illina}, year = {2021}, booktitle = {{Interspeech 2021}}, pages ...
We describe the LORIA-Inria-MULTISPEECH system submitted to the Oriental Language Recognition AP20-OLR Challenge. This system has been specifically designed to be robust to unknown conditions: channel mismatch (task 1) and noisy conditions (task 3). Three sets of studies have been carried out for elaborating the system...
null
null
kong21b_interspeech
Dynamic Multi-Scale Convolution for Dialect Identification
[ "Tianlong Kong", "Shouyi Yin", "Dawei Zhang", "Wang Geng", "Xin Wang", "Dandan Song", "Jinwen Huang", "Huiyu Shi", "Xiaorui Wang" ]
https://www.isca-archive.org/interspeech_2021/kong21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/kong21b_interspeech.pdf
10.21437/Interspeech.2021-56
3261-3265
@inproceedings{kong21b_interspeech, title = {{Dynamic Multi-Scale Convolution for Dialect Identification}}, author = {Tianlong Kong and Shouyi Yin and Dawei Zhang and Wang Geng and Xin Wang and Dandan Song and Jinwen Huang and Huiyu Shi and Xiaorui Wang}, year = {2021}, booktitle = {{Interspeech 202...
Time Delay Neural Networks (TDNN)-based methods are widely used in dialect identification. However, in previous work with TDNN application, subtle variant is being neglected in different feature scales. To address this issue, we propose a new architecture, named dynamic multi-scale convolution, which consists of dynami...
2108.07787
title_snapshot
wang21z_interspeech
An End-to-End Dialect Identification System with Transfer Learning from a Multilingual Automatic Speech Recognition Model
[ "Ding Wang", "Shuaishuai Ye", "Xinhui Hu", "Sheng Li", "Xinkang Xu" ]
https://www.isca-archive.org/interspeech_2021/wang21z_interspeech.html
https://www.isca-archive.org/interspeech_2021/wang21z_interspeech.pdf
10.21437/Interspeech.2021-374
3266-3270
@inproceedings{wang21z_interspeech, title = {{An End-to-End Dialect Identification System with Transfer Learning from a Multilingual Automatic Speech Recognition Model}}, author = {Ding Wang and Shuaishuai Ye and Xinhui Hu and Sheng Li and Xinkang Xu}, year = {2021}, booktitle = {{Interspeech 2021}}...
In this paper, we propose an end-to-end (E2E) dialect identification system trained using transfer learning from a multilingual automatic speech recognition (ASR) model. This is also an extension of our submitted system to the Oriental Language Recognition Challenge 2020 (AP20-OLR). We verified its applicability using ...
null
null
yu21b_interspeech
Language Recognition Based on Unsupervised Pretrained Models
[ "Haibin Yu", "Jing Zhao", "Song Yang", "Zhongqin Wu", "Yuting Nie", "Wei-Qiang Zhang" ]
https://www.isca-archive.org/interspeech_2021/yu21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/yu21b_interspeech.pdf
10.21437/Interspeech.2021-807
3271-3275
@inproceedings{yu21b_interspeech, title = {{Language Recognition Based on Unsupervised Pretrained Models}}, author = {Haibin Yu and Jing Zhao and Song Yang and Zhongqin Wu and Yuting Nie and Wei-Qiang Zhang}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3271--3275}, doi = ...
Unsupervised pretrained models have been proven to rival or even outperform supervised systems in various speech recognition tasks. However, their performance for language recognition is still left to be explored. In this paper, we construct several language recognition systems based on existing unsupervised pretrainin...
null
null
li21i_interspeech
Additive Phoneme-Aware Margin Softmax Loss for Language Recognition
[ "Zheng Li", "Yan Liu", "Lin Li", "Qingyang Hong" ]
https://www.isca-archive.org/interspeech_2021/li21i_interspeech.html
https://www.isca-archive.org/interspeech_2021/li21i_interspeech.pdf
10.21437/Interspeech.2021-1167
3276-3280
@inproceedings{li21i_interspeech, title = {{Additive Phoneme-Aware Margin Softmax Loss for Language Recognition}}, author = {Zheng Li and Yan Liu and Lin Li and Qingyang Hong}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3276--3280}, doi = {10.21437/Interspeech.2021-1167}...
This paper proposes an additive phoneme-aware margin softmax (APM-Softmax) loss to train the multi-task learning network with phonetic information for language recognition. In additive margin softmax (AM-Softmax) loss, the margin is set as a constant during the entire training for all training samples, and that is a su...
2106.12851
title_snapshot
jahchan21_interspeech
Towards an Accent-Robust Approach for ATC Communications Transcription
[ "Nataly Jahchan", "Florentin Barbier", "Ariyanidevi Dharma Gita", "Khaled Khelif", "Estelle Delpech" ]
https://www.isca-archive.org/interspeech_2021/jahchan21_interspeech.html
https://www.isca-archive.org/interspeech_2021/jahchan21_interspeech.pdf
10.21437/Interspeech.2021-333
3281-3285
@inproceedings{jahchan21_interspeech, title = {{Towards an Accent-Robust Approach for ATC Communications Transcription}}, author = {Nataly Jahchan and Florentin Barbier and Ariyanidevi Dharma Gita and Khaled Khelif and Estelle Delpech}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = ...
Air Traffic Control (ATC) communications are a typical example where Automatic Speech Recognition could face various challenges: audio data are quite noisy due to the characteristics of capturing mechanisms. All speakers involved use a specific English-based phraseology and a significant number of pilots and controller...
null
null
szoke21_interspeech
Detecting English Speech in the Air Traffic Control Voice Communication
[ "Igor Szöke", "Santosh Kesiraju", "Ondřej Novotný", "Martin Kocour", "Karel Veselý", "Jan Černocký" ]
https://www.isca-archive.org/interspeech_2021/szoke21_interspeech.html
https://www.isca-archive.org/interspeech_2021/szoke21_interspeech.pdf
10.21437/Interspeech.2021-1033
3286-3290
@inproceedings{szoke21_interspeech, title = {{Detecting English Speech in the Air Traffic Control Voice Communication}}, author = {Igor Szöke and Santosh Kesiraju and Ondřej Novotný and Martin Kocour and Karel Veselý and Jan Černocký}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {...
Developing in-cockpit voice enabled applications require a real-world dataset with labels and annotations. We launched a community platform for collecting the Air-Traffic Control (ATC) speech, world-wide in the ATCO 2 project. Filtering out non-English speech is one of the main components in the data processing pipelin...
2104.02332
title_snapshot
ohneiser21_interspeech
Robust Command Recognition for Lithuanian Air Traffic Control Tower Utterances
[ "Oliver Ohneiser", "Seyyed Saeed Sarfjoo", "Hartmut Helmke", "Shruthi Shetty", "Petr Motlicek", "Matthias Kleinert", "Heiko Ehr", "Šarūnas Murauskas" ]
https://www.isca-archive.org/interspeech_2021/ohneiser21_interspeech.html
https://www.isca-archive.org/interspeech_2021/ohneiser21_interspeech.pdf
10.21437/Interspeech.2021-935
3291-3295
@inproceedings{ohneiser21_interspeech, title = {{Robust Command Recognition for Lithuanian Air Traffic Control Tower Utterances}}, author = {Oliver Ohneiser and Seyyed Saeed Sarfjoo and Hartmut Helmke and Shruthi Shetty and Petr Motlicek and Matthias Kleinert and Heiko Ehr and Šarūnas Murauskas}, year ...
The maturity of automatic speech recognition (ASR) systems at controller working positions is currently a highly relevant technological topic in air traffic control (ATC). However, ATC service providers are less interested in pure word error rate (WER). They want to see benefits of ASR applications for ATC. Such applic...
null
null
zuluagagomez21_interspeech
Contextual Semi-Supervised Learning: An Approach to Leverage Air-Surveillance and Untranscribed ATC Data in ASR Systems
[ "Juan Zuluaga-Gomez", "Iuliia Nigmatulina", "Amrutha Prasad", "Petr Motlicek", "Karel Veselý", "Martin Kocour", "Igor Szöke" ]
https://www.isca-archive.org/interspeech_2021/zuluagagomez21_interspeech.html
https://www.isca-archive.org/interspeech_2021/zuluagagomez21_interspeech.pdf
10.21437/Interspeech.2021-1373
3296-3300
@inproceedings{zuluagagomez21_interspeech, title = {{Contextual Semi-Supervised Learning: An Approach to Leverage Air-Surveillance and Untranscribed ATC Data in ASR Systems}}, author = {Juan Zuluaga-Gomez and Iuliia Nigmatulina and Amrutha Prasad and Petr Motlicek and Karel Veselý and Martin Kocour and Igor ...
Air traffic management and specifically air-traffic control (ATC) rely mostly on voice communications between Air Traffic Controllers (ATCos) and pilots. In most cases, these voice communications follow a well-defined grammar that could be leveraged in Automatic Speech Recognition (ASR) technologies. The callsign used ...
2104.03643
title_snapshot
kocour21_interspeech
Boosting of Contextual Information in ASR for Air-Traffic Call-Sign Recognition
[ "Martin Kocour", "Karel Veselý", "Alexander Blatt", "Juan Zuluaga Gomez", "Igor Szöke", "Jan Černocký", "Dietrich Klakow", "Petr Motlicek" ]
https://www.isca-archive.org/interspeech_2021/kocour21_interspeech.html
https://www.isca-archive.org/interspeech_2021/kocour21_interspeech.pdf
10.21437/Interspeech.2021-1619
3301-3305
@inproceedings{kocour21_interspeech, title = {{Boosting of Contextual Information in ASR for Air-Traffic Call-Sign Recognition}}, author = {Martin Kocour and Karel Veselý and Alexander Blatt and Juan Zuluaga Gomez and Igor Szöke and Jan Černocký and Dietrich Klakow and Petr Motlicek}, year = {2021}, ...
Contextual adaptation of ASR can be very beneficial for multi-accent and often noisy Air-Traffic Control (ATC) speech. Our focus is call-sign recognition, which can be used to track conversations of ATC operators with individual airplanes. We developed a two-stage boosting strategy, consisting of HCLG boosting and Latt...
null
null
elie21_interspeech
Modeling the Effect of Military Oxygen Masks on Speech Characteristics
[ "Benjamin Elie", "Jodie Gauvain", "Jean-Luc Gauvain", "Lori Lamel" ]
https://www.isca-archive.org/interspeech_2021/elie21_interspeech.html
https://www.isca-archive.org/interspeech_2021/elie21_interspeech.pdf
10.21437/Interspeech.2021-1650
3306-3310
@inproceedings{elie21_interspeech, title = {{Modeling the Effect of Military Oxygen Masks on Speech Characteristics}}, author = {Benjamin Elie and Jodie Gauvain and Jean-Luc Gauvain and Lori Lamel}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3306--3310}, doi = {10.21437/...
Wearing an oxygen mask changes the speech production of speakers. It indeed modifies the vocal apparatus and perturbs the articulatory movements of the speaker. This paper studies the impact of the oxygen mask of military aircraft pilots on formant trajectories, both dynamically (variations of the formants at a utteran...
null
null
ribeiro21b_interspeech
Towards the Prediction of the Vocal Tract Shape from the Sequence of Phonemes to be Articulated
[ "Vinicius Ribeiro", "Karyna Isaieva", "Justine Leclere", "Pierre-André Vuissoz", "Yves Laprie" ]
https://www.isca-archive.org/interspeech_2021/ribeiro21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/ribeiro21b_interspeech.pdf
10.21437/Interspeech.2021-184
3325-3329
@inproceedings{ribeiro21b_interspeech, title = {{Towards the Prediction of the Vocal Tract Shape from the Sequence of Phonemes to be Articulated}}, author = {Vinicius Ribeiro and Karyna Isaieva and Justine Leclere and Pierre-André Vuissoz and Yves Laprie}, year = {2021}, booktitle = {{Interspeech 20...
In this work, we address the prediction of speech articulators’ temporal geometric position from the sequence of phonemes to be articulated. We start from a set of real-time MRI sequences uttered by a female French speaker. The contours of five articulators were tracked automatically in each of the frames in the MRI vi...
null
null
blandin21_interspeech
Comparison of the Finite Element Method, the Multimodal Method and the Transmission-Line Model for the Computation of Vocal Tract Transfer Functions
[ "Rémi Blandin", "Marc Arnela", "Simon Félix", "Jean-Baptiste Doc", "Peter Birkholz" ]
https://www.isca-archive.org/interspeech_2021/blandin21_interspeech.html
https://www.isca-archive.org/interspeech_2021/blandin21_interspeech.pdf
10.21437/Interspeech.2021-975
3330-3334
@inproceedings{blandin21_interspeech, title = {{Comparison of the Finite Element Method, the Multimodal Method and the Transmission-Line Model for the Computation of Vocal Tract Transfer Functions}}, author = {Rémi Blandin and Marc Arnela and Simon Félix and Jean-Baptiste Doc and Peter Birkholz}, year ...
The acoustic properties of vocal tract are usually characterized by its transfer function from the input acoustic volume flow at the glottis to the radiated acoustic pressure. These transfer functions can be computed with acoustic models. Three-dimensional acoustic simulation are used to take into account accurately th...
null
null
wagner21b_interspeech
Effects of Time Pressure and Spontaneity on Phonotactic Innovations in German Dialogues
[ "Petra Wagner", "Sina Zarrieß", "Joana Cholin" ]
https://www.isca-archive.org/interspeech_2021/wagner21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/wagner21b_interspeech.pdf
10.21437/Interspeech.2021-1539
3335-3339
@inproceedings{wagner21b_interspeech, title = {{Effects of Time Pressure and Spontaneity on Phonotactic Innovations in German Dialogues}}, author = {Petra Wagner and Sina Zarrieß and Joana Cholin}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3335--3339}, doi = {10.21437/I...
Speech variation is often explained by speakers’ balancing of production constraints (favoring phonetic reduction of high frequency, expected items) and listener orientation (favoring more canonical productions for low frequency, unexpected items). Less well understood are processes involving a structural reorganizatio...
null
null
medina21_interspeech
Importance of Parasagittal Sensor Information in Tongue Motion Capture Through a Diphonic Analysis
[ "Salvador Medina", "Sarah Taylor", "Mark Tiede", "Alexander Hauptmann", "Iain Matthews" ]
https://www.isca-archive.org/interspeech_2021/medina21_interspeech.html
https://www.isca-archive.org/interspeech_2021/medina21_interspeech.pdf
10.21437/Interspeech.2021-1732
3340-3344
@inproceedings{medina21_interspeech, title = {{Importance of Parasagittal Sensor Information in Tongue Motion Capture Through a Diphonic Analysis}}, author = {Salvador Medina and Sarah Taylor and Mark Tiede and Alexander Hauptmann and Iain Matthews}, year = {2021}, booktitle = {{Interspeech 2021}}, ...
Our study examines the information obtained by adding two parasagittal sensors to the standard midsagittal configuration of an Electromagnetic Articulography (EMA) observation of lingual articulation. In this work, we present a large and phonetically balanced corpus obtained from an EMA recording session of a single En...
null
null
georges21_interspeech
Learning Robust Speech Representation with an Articulatory-Regularized Variational Autoencoder
[ "Marc-Antoine Georges", "Laurent Girin", "Jean-Luc Schwartz", "Thomas Hueber" ]
https://www.isca-archive.org/interspeech_2021/georges21_interspeech.html
https://www.isca-archive.org/interspeech_2021/georges21_interspeech.pdf
10.21437/Interspeech.2021-1604
3345-3349
@inproceedings{georges21_interspeech, title = {{Learning Robust Speech Representation with an Articulatory-Regularized Variational Autoencoder}}, author = {Marc-Antoine Georges and Laurent Girin and Jean-Luc Schwartz and Thomas Hueber}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = ...
It is increasingly considered that human speech perception and production both rely on articulatory representations. In this paper, we investigate whether this type of representation could improve the performances of a deep generative model (here a variational autoencoder) trained to encode and decode acoustic speech f...
2104.03204
title_snapshot
weston21_interspeech
Changes in Glottal Source Parameter Values with Light to Moderate Physical Load
[ "Heather Weston", "Laura L. Koenig", "Susanne Fuchs" ]
https://www.isca-archive.org/interspeech_2021/weston21_interspeech.html
https://www.isca-archive.org/interspeech_2021/weston21_interspeech.pdf
10.21437/Interspeech.2021-1881
3350-3354
@inproceedings{weston21_interspeech, title = {{Changes in Glottal Source Parameter Values with Light to Moderate Physical Load}}, author = {Heather Weston and Laura L. Koenig and Susanne Fuchs}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3350--3354}, doi = {10.21437/Inte...
Engaging in everyday physical activities, like walking, initiates physiological processes that also affect parts of the body used for speech. However, it is currently unclear to what extent such activities affect phonatory processes, and in turn, the voice. The present exploratory study investigates how selected glotta...
null
null
vali21_interspeech
End-to-End Optimized Multi-Stage Vector Quantization of Spectral Envelopes for Speech and Audio Coding
[ "Mohammad Hassan Vali", "Tom Bäckström" ]
https://www.isca-archive.org/interspeech_2021/vali21_interspeech.html
https://www.isca-archive.org/interspeech_2021/vali21_interspeech.pdf
10.21437/Interspeech.2021-867
3355-3359
@inproceedings{vali21_interspeech, title = {{End-to-End Optimized Multi-Stage Vector Quantization of Spectral Envelopes for Speech and Audio Coding}}, author = {Mohammad Hassan Vali and Tom Bäckström}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3355--3359}, doi = {10.214...
Spectral envelope modeling is an instrumental part of speech and audio codecs, which can be used to enable efficient entropy coding of spectral components. Overall optimization of codecs, including envelope models, has however been difficult due to the complicated interactions between different modules of the codec. In...
null
null
nareddula21_interspeech
Fusion-Net: Time-Frequency Information Fusion Y-Network for Speech Enhancement
[ "Santhan Kumar Reddy Nareddula", "Subrahmanyam Gorthi", "Rama Krishna Sai S. Gorthi" ]
https://www.isca-archive.org/interspeech_2021/nareddula21_interspeech.html
https://www.isca-archive.org/interspeech_2021/nareddula21_interspeech.pdf
10.21437/Interspeech.2021-1184
3360-3364
@inproceedings{nareddula21_interspeech, title = {{Fusion-Net: Time-Frequency Information Fusion Y-Network for Speech Enhancement}}, author = {Santhan Kumar Reddy Nareddula and Subrahmanyam Gorthi and Rama Krishna Sai S. Gorthi}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3360--3...
This paper proposes a deep learning-based densely connected Y-Net as an effective network architecture for the fusion of time and frequency domain loss functions for speech enhancement. The proposed architecture performs speech enhancement in the time domain while fusing information from the frequency domain. Y-network...
null
null
marcinek21_interspeech
N-MTTL SI Model: Non-Intrusive Multi-Task Transfer Learning-Based Speech Intelligibility Prediction Model with Scenery Classification
[ "Ľuboš Marcinek", "Michael Stone", "Rebecca Millman", "Patrick Gaydecki" ]
https://www.isca-archive.org/interspeech_2021/marcinek21_interspeech.html
https://www.isca-archive.org/interspeech_2021/marcinek21_interspeech.pdf
10.21437/Interspeech.2021-1878
3365-3369
@inproceedings{marcinek21_interspeech, title = {{N-MTTL SI Model: Non-Intrusive Multi-Task Transfer Learning-Based Speech Intelligibility Prediction Model with Scenery Classification}}, author = {Ľuboš Marcinek and Michael Stone and Rebecca Millman and Patrick Gaydecki}, year = {2021}, booktitle = {...
The application of speech enhancement algorithms for hearing aids may not always be beneficial to increasing speech intelligibility. Therefore, a prior environment classification could be important. However, previous speech intelligibility models do not provide any additional information regarding the reason for a decr...
null
null
xia21b_interspeech
Temporal Context in Speech Emotion Recognition
[ "Yangyang Xia", "Li-Wei Chen", "Alexander Rudnicky", "Richard M. Stern" ]
https://www.isca-archive.org/interspeech_2021/xia21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/xia21b_interspeech.pdf
10.21437/Interspeech.2021-1840
3370-3374
@inproceedings{xia21b_interspeech, title = {{Temporal Context in Speech Emotion Recognition}}, author = {Yangyang Xia and Li-Wei Chen and Alexander Rudnicky and Richard M. Stern}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3370--3374}, doi = {10.21437/Interspeech.2021-18...
We investigate the importance of temporal context for speech emotion recognition (SER). Two SER systems trained on traditional and learned features, respectively, are developed to predict categorical labels of emotion. For traditional acoustical features, we study the combination of filterbank features and prosodic fea...
null
null
li21j_interspeech
Learning Fine-Grained Cross Modality Excitement for Speech Emotion Recognition
[ "Hang Li", "Wenbiao Ding", "Zhongqin Wu", "Zitao Liu" ]
https://www.isca-archive.org/interspeech_2021/li21j_interspeech.html
https://www.isca-archive.org/interspeech_2021/li21j_interspeech.pdf
10.21437/Interspeech.2021-158
3375-3379
@inproceedings{li21j_interspeech, title = {{Learning Fine-Grained Cross Modality Excitement for Speech Emotion Recognition}}, author = {Hang Li and Wenbiao Ding and Zhongqin Wu and Zitao Liu}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3375--3379}, doi = {10.21437/Inters...
Speech emotion recognition is a challenging task because the emotion expression is complex, multimodal and fine-grained. In this paper, we propose a novel multimodal deep learning approach to perform fine-grained emotion recognition from real-life speeches. We design a temporal alignment mean-max pooling mechanism to c...
2010.12733
title_snapshot
vaaras21_interspeech
Automatic Analysis of the Emotional Content of Speech in Daylong Child-Centered Recordings from a Neonatal Intensive Care Unit
[ "Einari Vaaras", "Sari Ahlqvist-Björkroth", "Konstantinos Drossos", "Okko Räsänen" ]
https://www.isca-archive.org/interspeech_2021/vaaras21_interspeech.html
https://www.isca-archive.org/interspeech_2021/vaaras21_interspeech.pdf
10.21437/Interspeech.2021-303
3380-3384
@inproceedings{vaaras21_interspeech, title = {{Automatic Analysis of the Emotional Content of Speech in Daylong Child-Centered Recordings from a Neonatal Intensive Care Unit}}, author = {Einari Vaaras and Sari Ahlqvist-Björkroth and Konstantinos Drossos and Okko Räsänen}, year = {2021}, booktitle = ...
Researchers have recently started to study how the emotional speech heard by young infants can affect their developmental outcomes. As a part of this research, hundreds of hours of daylong recordings from preterm infants’ audio environments were collected from two hospitals in Finland and Estonia in the context of so-c...
2106.09539
title_snapshot
qian21_interspeech
Multimodal Sentiment Analysis with Temporal Modality Attention
[ "Fan Qian", "Jiqing Han" ]
https://www.isca-archive.org/interspeech_2021/qian21_interspeech.html
https://www.isca-archive.org/interspeech_2021/qian21_interspeech.pdf
10.21437/Interspeech.2021-487
3385-3389
@inproceedings{qian21_interspeech, title = {{Multimodal Sentiment Analysis with Temporal Modality Attention}}, author = {Fan Qian and Jiqing Han}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3385--3389}, doi = {10.21437/Interspeech.2021-487}, issn = {2958-1796}, }
Multimodal sentiment analysis is an important research that involves integrating information from multiple modalities to identify a speaker underlying attitude. The core challenge is to model cross-modal interactions which span across both the different modalities and time. Although great progress has been made, the ex...
null
null
t21_interspeech
Stochastic Process Regression for Cross-Cultural Speech Emotion Recognition
[ "Mani Kumar T", "Enrique Sanchez", "Georgios Tzimiropoulos", "Timo Giesbrecht", "Michel Valstar" ]
https://www.isca-archive.org/interspeech_2021/t21_interspeech.html
https://www.isca-archive.org/interspeech_2021/t21_interspeech.pdf
10.21437/Interspeech.2021-610
3390-3394
@inproceedings{t21_interspeech, title = {{Stochastic Process Regression for Cross-Cultural Speech Emotion Recognition}}, author = {Mani Kumar T and Enrique Sanchez and Georgios Tzimiropoulos and Timo Giesbrecht and Michel Valstar}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3390...
In this work, we pose continuous apparent emotion recognition from speech as a problem of learning distributions of functions, and do so using Stochastic Processes Regression. We presume that the relation between speech signals and their corresponding emotion labels is governed by some underlying stochastic process, in...
null
null
li21k_interspeech
Acted vs. Improvised: Domain Adaptation for Elicitation Approaches in Audio-Visual Emotion Recognition
[ "Haoqi Li", "Yelin Kim", "Cheng-Hao Kuo", "Shrikanth S. Narayanan" ]
https://www.isca-archive.org/interspeech_2021/li21k_interspeech.html
https://www.isca-archive.org/interspeech_2021/li21k_interspeech.pdf
10.21437/Interspeech.2021-666
3395-3399
@inproceedings{li21k_interspeech, title = {{Acted vs. Improvised: Domain Adaptation for Elicitation Approaches in Audio-Visual Emotion Recognition}}, author = {Haoqi Li and Yelin Kim and Cheng-Hao Kuo and Shrikanth S. Narayanan}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3395--...
Key challenges in developing generalized automatic emotion recognition systems include scarcity of labeled data and lack of gold-standard references. Even for the cues that are labeled as the same emotion category, the variability of associated expressions can be high depending on the elicitation context e.g., emotion ...
2104.01978
title_snapshot
pepino21_interspeech
Emotion Recognition from Speech Using wav2vec 2.0 Embeddings
[ "Leonardo Pepino", "Pablo Riera", "Luciana Ferrer" ]
https://www.isca-archive.org/interspeech_2021/pepino21_interspeech.html
https://www.isca-archive.org/interspeech_2021/pepino21_interspeech.pdf
10.21437/Interspeech.2021-703
3400-3404
@inproceedings{pepino21_interspeech, title = {{Emotion Recognition from Speech Using wav2vec 2.0 Embeddings}}, author = {Leonardo Pepino and Pablo Riera and Luciana Ferrer}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3400--3404}, doi = {10.21437/Interspeech.2021-703}, ...
Emotion recognition datasets are relatively small, making the use of deep learning techniques challenging. In this work, we propose a transfer learning method for speech emotion recognition (SER) where features extracted from pre-trained wav2vec 2.0 models are used as input to shallow neural networks to recognize emoti...
2104.03502
title_snapshot
liu21k_interspeech
Graph Isomorphism Network for Speech Emotion Recognition
[ "Jiawang Liu", "Haoxiang Wang" ]
https://www.isca-archive.org/interspeech_2021/liu21k_interspeech.html
https://www.isca-archive.org/interspeech_2021/liu21k_interspeech.pdf
10.21437/Interspeech.2021-1154
3405-3409
@inproceedings{liu21k_interspeech, title = {{Graph Isomorphism Network for Speech Emotion Recognition}}, author = {Jiawang Liu and Haoxiang Wang}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3405--3409}, doi = {10.21437/Interspeech.2021-1154}, issn = {2958-1796}, }
Previous deep learning approaches such as Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) have been broadly used in speech emotion recognition (SER). In these approaches, speech signals are generally modeled in the Euclidean space. In this paper, a novel SER model (LSTM-GIN) is proposed, which appl...
null
null
kumawat21_interspeech
Applying TDNN Architectures for Analyzing Duration Dependencies on Speech Emotion Recognition
[ "Pooja Kumawat", "Aurobinda Routray" ]
https://www.isca-archive.org/interspeech_2021/kumawat21_interspeech.html
https://www.isca-archive.org/interspeech_2021/kumawat21_interspeech.pdf
10.21437/Interspeech.2021-2168
3410-3414
@inproceedings{kumawat21_interspeech, title = {{Applying TDNN Architectures for Analyzing Duration Dependencies on Speech Emotion Recognition}}, author = {Pooja Kumawat and Aurobinda Routray}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3410--3414}, doi = {10.21437/Inters...
We have analyzed the Time Delay Neural Network (TDNN) based architectures for speech emotion classification. TDNN models efficiently capture the temporal information and provide an utterance level prediction. Emotions are dynamic in nature and require temporal context for reliable prediction. In our work, we have appli...
null
null
keesing21_interspeech
Acoustic Features and Neural Representations for Categorical Emotion Recognition from Speech
[ "Aaron Keesing", "Yun Sing Koh", "Michael Witbrock" ]
https://www.isca-archive.org/interspeech_2021/keesing21_interspeech.html
https://www.isca-archive.org/interspeech_2021/keesing21_interspeech.pdf
10.21437/Interspeech.2021-2217
3415-3419
@inproceedings{keesing21_interspeech, title = {{Acoustic Features and Neural Representations for Categorical Emotion Recognition from Speech}}, author = {Aaron Keesing and Yun Sing Koh and Michael Witbrock}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3415--3419}, doi = {...
Many features have been proposed for use in speech emotion recognition, from signal processing features to bag-of-audio-words (BoAW) models to abstract neural representations. Some of these feature types have not been directly compared across a large number of speech corpora to determine performance differences. We pro...
null
null
shon21_interspeech
Leveraging Pre-Trained Language Model for Speech Sentiment Analysis
[ "Suwon Shon", "Pablo Brusco", "Jing Pan", "Kyu J. Han", "Shinji Watanabe" ]
https://www.isca-archive.org/interspeech_2021/shon21_interspeech.html
https://www.isca-archive.org/interspeech_2021/shon21_interspeech.pdf
10.21437/Interspeech.2021-1723
3420-3424
@inproceedings{shon21_interspeech, title = {{Leveraging Pre-Trained Language Model for Speech Sentiment Analysis}}, author = {Suwon Shon and Pablo Brusco and Jing Pan and Kyu J. Han and Shinji Watanabe}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3420--3424}, doi = {10.2...
In this paper, we explore the use of pre-trained language models to learn sentiment information of written texts for speech sentiment analysis. First, we investigate how useful a pre-trained language model would be in a 2-step pipeline approach employing Automatic Speech Recognition (ASR) and transcripts-based sentimen...
2106.06598
title_snapshot
hou21b_interspeech
Cross-Domain Speech Recognition with Unsupervised Character-Level Distribution Matching
[ "Wenxin Hou", "Jindong Wang", "Xu Tan", "Tao Qin", "Takahiro Shinozaki" ]
https://www.isca-archive.org/interspeech_2021/hou21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/hou21b_interspeech.pdf
10.21437/Interspeech.2021-57
3425-3429
@inproceedings{hou21b_interspeech, title = {{Cross-Domain Speech Recognition with Unsupervised Character-Level Distribution Matching}}, author = {Wenxin Hou and Jindong Wang and Xu Tan and Tao Qin and Takahiro Shinozaki}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3425--3429}, ...
End-to-end automatic speech recognition (ASR) can achieve promising performance with large-scale training data. However, it is known that domain mismatch between training and testing data often leads to a degradation of recognition accuracy. In this work, we focus on the unsupervised domain adaptation for ASR and propo...
2104.07491
title_snapshot
kanda21_interspeech
Large-Scale Pre-Training of End-to-End Multi-Talker ASR for Meeting Transcription with Single Distant Microphone
[ "Naoyuki Kanda", "Guoli Ye", "Yu Wu", "Yashesh Gaur", "Xiaofei Wang", "Zhong Meng", "Zhuo Chen", "Takuya Yoshioka" ]
https://www.isca-archive.org/interspeech_2021/kanda21_interspeech.html
https://www.isca-archive.org/interspeech_2021/kanda21_interspeech.pdf
10.21437/Interspeech.2021-102
3430-3434
@inproceedings{kanda21_interspeech, title = {{Large-Scale Pre-Training of End-to-End Multi-Talker ASR for Meeting Transcription with Single Distant Microphone}}, author = {Naoyuki Kanda and Guoli Ye and Yu Wu and Yashesh Gaur and Xiaofei Wang and Zhong Meng and Zhuo Chen and Takuya Yoshioka}, year = {...
Transcribing meetings containing overlapped speech with only a single distant microphone (SDM) has been one of the most challenging problems for automatic speech recognition (ASR). While various approaches have been proposed, all previous studies on the monaural overlapped speech recognition problem were based on eithe...
2103.16776
title_snapshot
lu21b_interspeech
On Minimum Word Error Rate Training of the Hybrid Autoregressive Transducer
[ "Liang Lu", "Zhong Meng", "Naoyuki Kanda", "Jinyu Li", "Yifan Gong" ]
https://www.isca-archive.org/interspeech_2021/lu21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/lu21b_interspeech.pdf
10.21437/Interspeech.2021-161
3435-3439
@inproceedings{lu21b_interspeech, title = {{On Minimum Word Error Rate Training of the Hybrid Autoregressive Transducer}}, author = {Liang Lu and Zhong Meng and Naoyuki Kanda and Jinyu Li and Yifan Gong}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3435--3439}, doi = {10....
Hybrid Autoregressive Transducer (HAT) is a recently proposed end-to-end acoustic model that extends the standard Recurrent Neural Network Transducer (RNN-T) for the purpose of the external language model (LM) fusion. In HAT, the blank probability and the label probability are estimated using two separate probability d...
2010.12673
title_snapshot
kim21j_interspeech
Reducing Streaming ASR Model Delay with Self Alignment
[ "Jaeyoung Kim", "Han Lu", "Anshuman Tripathi", "Qian Zhang", "Hasim Sak" ]
https://www.isca-archive.org/interspeech_2021/kim21j_interspeech.html
https://www.isca-archive.org/interspeech_2021/kim21j_interspeech.pdf
10.21437/Interspeech.2021-322
3440-3444
@inproceedings{kim21j_interspeech, title = {{Reducing Streaming ASR Model Delay with Self Alignment}}, author = {Jaeyoung Kim and Han Lu and Anshuman Tripathi and Qian Zhang and Hasim Sak}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3440--3444}, doi = {10.21437/Interspee...
Reducing prediction delay for streaming end-to-end ASR models with minimal performance regression is a challenging problem. Constrained alignment is a well-known existing approach that penalizes predicted word boundaries using external low-latency acoustic models. On the contrary, recently proposed FastEmit is a sequen...
2105.05005
title_snapshot
diwan21b_interspeech
Reduce and Reconstruct: ASR for Low-Resource Phonetic Languages
[ "Anuj Diwan", "Preethi Jyothi" ]
https://www.isca-archive.org/interspeech_2021/diwan21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/diwan21b_interspeech.pdf
10.21437/Interspeech.2021-644
3445-3449
@inproceedings{diwan21b_interspeech, title = {{Reduce and Reconstruct: ASR for Low-Resource Phonetic Languages}}, author = {Anuj Diwan and Preethi Jyothi}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3445--3449}, doi = {10.21437/Interspeech.2021-644}, issn = {2958-...
This work presents a seemingly simple but effective technique to improve low-resource ASR systems for phonetic languages. By identifying sets of acoustically similar graphemes in these languages, we first reduce the output alphabet of the ASR system using linguistically meaningful reductions and then reconstruct the or...
2010.09322
title_snapshot
fukuda21_interspeech
Knowledge Distillation Based Training of Universal ASR Source Models for Cross-Lingual Transfer
[ "Takashi Fukuda", "Samuel Thomas" ]
https://www.isca-archive.org/interspeech_2021/fukuda21_interspeech.html
https://www.isca-archive.org/interspeech_2021/fukuda21_interspeech.pdf
10.21437/Interspeech.2021-796
3450-3454
@inproceedings{fukuda21_interspeech, title = {{Knowledge Distillation Based Training of Universal ASR Source Models for Cross-Lingual Transfer}}, author = {Takashi Fukuda and Samuel Thomas}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3450--3454}, doi = {10.21437/Interspe...
In this paper we introduce a novel knowledge distillation based framework for training universal source models. In our proposed approach for automatic speech recognition (ASR), multilingual source models are first trained using multiple language-dependent resources before being used to initialize language specific targ...
null
null
ray21_interspeech
Listen with Intent: Improving Speech Recognition with Audio-to-Intent Front-End
[ "Swayambhu Nath Ray", "Minhua Wu", "Anirudh Raju", "Pegah Ghahremani", "Raghavendra Bilgi", "Milind Rao", "Harish Arsikere", "Ariya Rastrow", "Andreas Stolcke", "Jasha Droppo" ]
https://www.isca-archive.org/interspeech_2021/ray21_interspeech.html
https://www.isca-archive.org/interspeech_2021/ray21_interspeech.pdf
10.21437/Interspeech.2021-836
3455-3459
@inproceedings{ray21_interspeech, title = {{Listen with Intent: Improving Speech Recognition with Audio-to-Intent Front-End}}, author = {Swayambhu Nath Ray and Minhua Wu and Anirudh Raju and Pegah Ghahremani and Raghavendra Bilgi and Milind Rao and Harish Arsikere and Ariya Rastrow and Andreas Stolcke and Ja...
Comprehending the overall intent of an utterance helps a listener recognize the individual words spoken. Inspired by this fact, we perform a novel study of the impact of explicitly incorporating intent representations as additional information to improve a recurrent neural network-transducer (RNN-T) based automatic spe...
2105.07071
title_snapshot
lu21c_interspeech
Exploring Targeted Universal Adversarial Perturbations to End-to-End ASR Models
[ "Zhiyun Lu", "Wei Han", "Yu Zhang", "Liangliang Cao" ]
https://www.isca-archive.org/interspeech_2021/lu21c_interspeech.html
https://www.isca-archive.org/interspeech_2021/lu21c_interspeech.pdf
10.21437/Interspeech.2021-1668
3460-3464
@inproceedings{lu21c_interspeech, title = {{Exploring Targeted Universal Adversarial Perturbations to End-to-End ASR Models}}, author = {Zhiyun Lu and Wei Han and Yu Zhang and Liangliang Cao}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3460--3464}, doi = {10.21437/Inters...
Although end-to-end automatic speech recognition (e2e ASR) models are widely deployed in many applications, there have been very few studies to understand models’ robustness against adversarial perturbations. In this paper, we explore whether a targeted universal perturbation vector exists for e2e ASR models. Our goal ...
2104.02757
title_snapshot
delrio21_interspeech
Earnings-21: A Practical Benchmark for ASR in the Wild
[ "Miguel Del Rio", "Natalie Delworth", "Ryan Westerman", "Michelle Huang", "Nishchal Bhandari", "Joseph Palakapilly", "Quinten McNamara", "Joshua Dong", "Piotr Żelasko", "Miguel Jetté" ]
https://www.isca-archive.org/interspeech_2021/delrio21_interspeech.html
https://www.isca-archive.org/interspeech_2021/delrio21_interspeech.pdf
10.21437/Interspeech.2021-1915
3465-3469
@inproceedings{delrio21_interspeech, title = {{Earnings-21: A Practical Benchmark for ASR in the Wild}}, author = {Miguel {Del Rio} and Natalie Delworth and Ryan Westerman and Michelle Huang and Nishchal Bhandari and Joseph Palakapilly and Quinten McNamara and Joshua Dong and Piotr Żelasko and Miguel Jetté},...
Commonly used speech corpora inadequately challenge academic and commercial ASR systems. In particular, speech corpora lack metadata needed for detailed analysis and WER measurement. In response, we present Earnings-21 , a 39-hour corpus of earnings calls containing entity-dense speech from nine different financial sec...
2104.11348
title_snapshot
sun21c_interspeech
Improving Multilingual Transformer Transducer Models by Reducing Language Confusions
[ "Eric Sun", "Jinyu Li", "Zhong Meng", "Yu Wu", "Jian Xue", "Shujie Liu", "Yifan Gong" ]
https://www.isca-archive.org/interspeech_2021/sun21c_interspeech.html
https://www.isca-archive.org/interspeech_2021/sun21c_interspeech.pdf
10.21437/Interspeech.2021-1949
3470-3474
@inproceedings{sun21c_interspeech, title = {{Improving Multilingual Transformer Transducer Models by Reducing Language Confusions}}, author = {Eric Sun and Jinyu Li and Zhong Meng and Yu Wu and Jian Xue and Shujie Liu and Yifan Gong}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3...
In end-to-end multilingual speech recognition, the hypotheses in one language could include word tokens from other languages. Language confusions happen even more frequently when language identifier (LID) is not present during inference. In this paper, we explore to reduce language confusions without using LID in model...
null
null
ali21b_interspeech
Arabic Code-Switching Speech Recognition Using Monolingual Data
[ "Ahmed Ali", "Shammur Absar Chowdhury", "Amir Hussein", "Yasser Hifny" ]
https://www.isca-archive.org/interspeech_2021/ali21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/ali21b_interspeech.pdf
10.21437/Interspeech.2021-2231
3475-3479
@inproceedings{ali21b_interspeech, title = {{Arabic Code-Switching Speech Recognition Using Monolingual Data}}, author = {Ahmed Ali and Shammur Absar Chowdhury and Amir Hussein and Yasser Hifny}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3475--3479}, doi = {10.21437/Int...
Code-switching in automatic speech recognition (ASR) is an important challenge due to globalization. Recent research in multilingual ASR shows potential improvement over monolingual systems. We study key issues related to multilingual modeling for ASR through a series of large-scale ASR experiments. Our innovative fram...
2107.01573
title_snapshot
eisenberg21_interspeech
Online Blind Audio Source Separation Using Recursive Expectation-Maximization
[ "Aviad Eisenberg", "Boaz Schwartz", "Sharon Gannot" ]
https://www.isca-archive.org/interspeech_2021/eisenberg21_interspeech.html
https://www.isca-archive.org/interspeech_2021/eisenberg21_interspeech.pdf
10.21437/Interspeech.2021-662
3480-3484
@inproceedings{eisenberg21_interspeech, title = {{Online Blind Audio Source Separation Using Recursive Expectation-Maximization}}, author = {Aviad Eisenberg and Boaz Schwartz and Sharon Gannot}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3480--3484}, doi = {10.21437/Inte...
The challenging problem of online multi-microphone blind audio source separation (BASS) in noisy environment is addressed in this paper. We present a sequential, non-iterative, algorithm based on the recursive EM (REM) framework. In the proposed algorithm, the compete-data, which constitutes the separated sources and r...
null
null
luo21e_interspeech
Empirical Analysis of Generalized Iterative Speech Separation Networks
[ "Yi Luo", "Cong Han", "Nima Mesgarani" ]
https://www.isca-archive.org/interspeech_2021/luo21e_interspeech.html
https://www.isca-archive.org/interspeech_2021/luo21e_interspeech.pdf
10.21437/Interspeech.2021-1161
3485-3489
@inproceedings{luo21e_interspeech, title = {{Empirical Analysis of Generalized Iterative Speech Separation Networks}}, author = {Yi Luo and Cong Han and Nima Mesgarani}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3485--3489}, doi = {10.21437/Interspeech.2021-1161}, iss...
Although most existing speech separation networks are designed as a one-pass pipeline where the sources are directly estimated from the mixture, multi-pass or iterative pipelines have been shown to be effective by designing multiple rounds of separation and utilizing separation outputs from a previous iteration as addi...
null
null
neumann21_interspeech
Graph-PIT: Generalized Permutation Invariant Training for Continuous Separation of Arbitrary Numbers of Speakers
[ "Thilo von Neumann", "Keisuke Kinoshita", "Christoph Boeddeker", "Marc Delcroix", "Reinhold Haeb-Umbach" ]
https://www.isca-archive.org/interspeech_2021/neumann21_interspeech.html
https://www.isca-archive.org/interspeech_2021/neumann21_interspeech.pdf
10.21437/Interspeech.2021-1177
3490-3494
@inproceedings{neumann21_interspeech, title = {{Graph-PIT: Generalized Permutation Invariant Training for Continuous Separation of Arbitrary Numbers of Speakers}}, author = {Thilo {von Neumann} and Keisuke Kinoshita and Christoph Boeddeker and Marc Delcroix and Reinhold Haeb-Umbach}, year = {2021}, ...
Automatic transcription of meetings requires handling of overlapped speech, which calls for continuous speech separation (CSS) systems. The uPIT criterion was proposed for utterance-level separation with neural networks and introduces the constraint that the total number of speakers must not exceed the number of output...
2107.14446
title_snapshot
zhang21v_interspeech
Teacher-Student MixIT for Unsupervised and Semi-Supervised Speech Separation
[ "Jisi Zhang", "Cătălin Zorilă", "Rama Doddipatla", "Jon Barker" ]
https://www.isca-archive.org/interspeech_2021/zhang21v_interspeech.html
https://www.isca-archive.org/interspeech_2021/zhang21v_interspeech.pdf
10.21437/Interspeech.2021-1243
3495-3499
@inproceedings{zhang21v_interspeech, title = {{Teacher-Student MixIT for Unsupervised and Semi-Supervised Speech Separation}}, author = {Jisi Zhang and Cătălin Zorilă and Rama Doddipatla and Jon Barker}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3495--3499}, doi = {10.2...
In this paper, we introduce a novel semi-supervised learning framework for end-to-end speech separation. The proposed method first uses mixtures of unseparated sources and the mixture invariant training (MixIT) criterion to train a teacher model. The teacher model then estimates separated sources that are used to train...
2106.07843
title_snapshot
delcroix21_interspeech
Few-Shot Learning of New Sound Classes for Target Sound Extraction
[ "Marc Delcroix", "Jorge Bennasar Vázquez", "Tsubasa Ochiai", "Keisuke Kinoshita", "Shoko Araki" ]
https://www.isca-archive.org/interspeech_2021/delcroix21_interspeech.html
https://www.isca-archive.org/interspeech_2021/delcroix21_interspeech.pdf
10.21437/Interspeech.2021-1369
3500-3504
@inproceedings{delcroix21_interspeech, title = {{Few-Shot Learning of New Sound Classes for Target Sound Extraction}}, author = {Marc Delcroix and Jorge Bennasar Vázquez and Tsubasa Ochiai and Keisuke Kinoshita and Shoko Araki}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3500--3...
Target sound extraction consists of extracting the sound of a target acoustic event (AE) class from a mixture of AE sounds. It can be realized using a neural network that extracts the target sound conditioned on a 1-hot vector that represents the desired AE class. With this approach, embedding vectors associated with t...
2106.07144
title_snapshot
han21e_interspeech
Binaural Speech Separation of Moving Speakers With Preserved Spatial Cues
[ "Cong Han", "Yi Luo", "Nima Mesgarani" ]
https://www.isca-archive.org/interspeech_2021/han21e_interspeech.html
https://www.isca-archive.org/interspeech_2021/han21e_interspeech.pdf
10.21437/Interspeech.2021-1372
3505-3509
@inproceedings{han21e_interspeech, title = {{Binaural Speech Separation of Moving Speakers With Preserved Spatial Cues}}, author = {Cong Han and Yi Luo and Nima Mesgarani}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3505--3509}, doi = {10.21437/Interspeech.2021-1372}, ...
Binaural speech separation algorithms designed for augmented hearing technologies need to both improve the signal-to-noise ratio of individual speakers and preserve their perceived location in space. The majority of binaural speech separation methods assume nonmoving speakers. As a result, their application to real-wor...
null
null
hu21_interspeech
AvaTr: One-Shot Speaker Extraction with Transformers
[ "Shell Xu Hu", "Md. Rifat Arefin", "Viet-Nhat Nguyen", "Alish Dipani", "Xaq Pitkow", "Andreas Savas Tolias" ]
https://www.isca-archive.org/interspeech_2021/hu21_interspeech.html
https://www.isca-archive.org/interspeech_2021/hu21_interspeech.pdf
10.21437/Interspeech.2021-1378
3510-3514
@inproceedings{hu21_interspeech, title = {{AvaTr: One-Shot Speaker Extraction with Transformers}}, author = {Shell Xu Hu and Md. Rifat Arefin and Viet-Nhat Nguyen and Alish Dipani and Xaq Pitkow and Andreas Savas Tolias}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3510--3514}, ...
To extract the voice of a target speaker when mixed with a variety of other sounds, such as white and ambient noises or the voices of interfering speakers, we extend the Transformer network [1] to attend the most relevant information with respect to the target speaker given the characteristics of his or her voices as a...
2105.00609
title_snapshot
sarkar21_interspeech
Vocal Harmony Separation Using Time-Domain Neural Networks
[ "Saurjya Sarkar", "Emmanouil Benetos", "Mark Sandler" ]
https://www.isca-archive.org/interspeech_2021/sarkar21_interspeech.html
https://www.isca-archive.org/interspeech_2021/sarkar21_interspeech.pdf
10.21437/Interspeech.2021-1531
3515-3519
@inproceedings{sarkar21_interspeech, title = {{Vocal Harmony Separation Using Time-Domain Neural Networks}}, author = {Saurjya Sarkar and Emmanouil Benetos and Mark Sandler}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3515--3519}, doi = {10.21437/Interspeech.2021-1531}, ...
Polyphonic vocal recordings are an inherently challenging source separation task due to the melodic structure of the vocal parts and unique timbre of its constituents. In this work we utilise a time-domain neural network architecture re-purposed from speech separation research and modify it to separate a capella mixtur...
null
null
maciejewski21_interspeech
Speaker Verification-Based Evaluation of Single-Channel Speech Separation
[ "Matthew Maciejewski", "Shinji Watanabe", "Sanjeev Khudanpur" ]
https://www.isca-archive.org/interspeech_2021/maciejewski21_interspeech.html
https://www.isca-archive.org/interspeech_2021/maciejewski21_interspeech.pdf
10.21437/Interspeech.2021-1924
3520-3524
@inproceedings{maciejewski21_interspeech, title = {{Speaker Verification-Based Evaluation of Single-Channel Speech Separation}}, author = {Matthew Maciejewski and Shinji Watanabe and Sanjeev Khudanpur}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3520--3524}, doi = {10.21...
Speech enhancement techniques typically focus on intrinsic metrics of signal quality. The overwhelming majority of deep learning-based single-channel speech separation studies, for instance, have relied on a single class of metrics to evaluate the systems by. These metrics, usually variants of Signal-to-Distortion Rati...
null
null
lan21_interspeech
Improved Speech Separation with Time-and-Frequency Cross-Domain Feature Selection
[ "Tian Lan", "Yuxin Qian", "Yilan Lyu", "Refuoe Mokhosi", "Wenxin Tai", "Qiao Liu" ]
https://www.isca-archive.org/interspeech_2021/lan21_interspeech.html
https://www.isca-archive.org/interspeech_2021/lan21_interspeech.pdf
10.21437/Interspeech.2021-2246
3525-3529
@inproceedings{lan21_interspeech, title = {{Improved Speech Separation with Time-and-Frequency Cross-Domain Feature Selection}}, author = {Tian Lan and Yuxin Qian and Yilan Lyu and Refuoe Mokhosi and Wenxin Tai and Qiao Liu}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3525--3529...
Most deep learning-based monaural speech separation models only use either spectrograms or time domain speech signal as the input feature. The recently proposed cross-domain network (CDNet) demonstrates that concatenated frequency domain and time domain features helps to reach better performance. Although concatenation...
null
null
deng21c_interspeech
Robust Speaker Extraction Network Based on Iterative Refined Adaptation
[ "Chengyun Deng", "Shiqian Ma", "Yongtao Sha", "Yi Zhang", "Hui Zhang", "Hui Song", "Fei Wang" ]
https://www.isca-archive.org/interspeech_2021/deng21c_interspeech.html
https://www.isca-archive.org/interspeech_2021/deng21c_interspeech.pdf
10.21437/Interspeech.2021-2250
3530-3534
@inproceedings{deng21c_interspeech, title = {{Robust Speaker Extraction Network Based on Iterative Refined Adaptation}}, author = {Chengyun Deng and Shiqian Ma and Yongtao Sha and Yi Zhang and Hui Zhang and Hui Song and Fei Wang}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3530-...
Speaker extraction aims to extract target speech signal from a multi-talker environment with interference speakers and surrounding noise, given a reference speech from target speaker. Most speaker extraction systems achieve satisfactory performance in the closed condition. Such systems suffer from performance degradati...
2011.02102
title_snapshot
wang21aa_interspeech
Neural Speaker Extraction with Speaker-Speech Cross-Attention Network
[ "Wupeng Wang", "Chenglin Xu", "Meng Ge", "Haizhou Li" ]
https://www.isca-archive.org/interspeech_2021/wang21aa_interspeech.html
https://www.isca-archive.org/interspeech_2021/wang21aa_interspeech.pdf
10.21437/Interspeech.2021-2260
3535-3539
@inproceedings{wang21aa_interspeech, title = {{Neural Speaker Extraction with Speaker-Speech Cross-Attention Network}}, author = {Wupeng Wang and Chenglin Xu and Meng Ge and Haizhou Li}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3535--3539}, doi = {10.21437/Interspeech....
In this paper, we propose a novel time-domain speaker-speech cross-attention network as a variant of SpEx [1] architecture, that features speaker-speech cross-attention. The speaker-speech cross-attention network consists of speech semantic layers that capture the high-level dependency of audio feature, and cross-atten...
null
null
rigal21_interspeech
Deep Audio-Visual Speech Separation Based on Facial Motion
[ "Rémi Rigal", "Jacques Chodorowski", "Benoît Zerr" ]
https://www.isca-archive.org/interspeech_2021/rigal21_interspeech.html
https://www.isca-archive.org/interspeech_2021/rigal21_interspeech.pdf
10.21437/Interspeech.2021-1560
3540-3544
@inproceedings{rigal21_interspeech, title = {{Deep Audio-Visual Speech Separation Based on Facial Motion}}, author = {Rémi Rigal and Jacques Chodorowski and Benoît Zerr}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {3540--3544}, doi = {10.21437/Interspeech.2021-1560}, is...
We present a deep neural network that relies on facial motion and time-domain audio for isolating speech signals from a mixture of speeches and background noises. Recent studies in deep learning-based audio-visual speech separation and speech enhancement have proven that leveraging visual information in addition to aud...
null
null