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
chanclu21_interspeech
Automatic Classification of Phonation Types in Spontaneous Speech: Towards a New Workflow for the Characterization of Speakers’ Voice Quality
[ "Anaïs Chanclu", "Imen Ben Amor", "Cédric Gendrot", "Emmanuel Ferragne", "Jean-François Bonastre" ]
https://www.isca-archive.org/interspeech_2021/chanclu21_interspeech.html
https://www.isca-archive.org/interspeech_2021/chanclu21_interspeech.pdf
10.21437/Interspeech.2021-1765
1015-1018
@inproceedings{chanclu21_interspeech, title = {{Automatic Classification of Phonation Types in Spontaneous Speech: Towards a New Workflow for the Characterization of Speakers’ Voice Quality}}, author = {Anaïs Chanclu and Imen Ben Amor and Cédric Gendrot and Emmanuel Ferragne and Jean-François Bonastre}, ye...
Voice quality is known to be an important factor for the characterization of a speaker’s voice, both in terms of physiological features (mainly laryngeal and supralaryngeal) and of the speaker’s habits (sociolinguistic factors). This paper is devoted to one of the main components of voice quality: phonation type. It pr...
null
null
son21_interspeech
Measuring Voice Quality Parameters After Speaker Pseudonymization
[ "Rob J.J.H. van Son" ]
https://www.isca-archive.org/interspeech_2021/son21_interspeech.html
https://www.isca-archive.org/interspeech_2021/son21_interspeech.pdf
10.21437/Interspeech.2021-26
1019-1023
@inproceedings{son21_interspeech, title = {{Measuring Voice Quality Parameters After Speaker Pseudonymization}}, author = {Rob J.J.H. van Son}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1019--1023}, doi = {10.21437/Interspeech.2021-26}, issn = {2958-1796}, }
Collecting and sharing speech resources is important for progress in speech science and technology. Often, speech resources cannot be shared because of concerns over the privacy of the speakers, e.g., minors or people with medical conditions. Current technologies for pseudonymizing speech have only been tested on “stan...
null
null
steinert21_interspeech
Audio-Visual Recognition of Emotional Engagement of People with Dementia
[ "Lars Steinert", "Felix Putze", "Dennis Küster", "Tanja Schultz" ]
https://www.isca-archive.org/interspeech_2021/steinert21_interspeech.html
https://www.isca-archive.org/interspeech_2021/steinert21_interspeech.pdf
10.21437/Interspeech.2021-567
1024-1028
@inproceedings{steinert21_interspeech, title = {{Audio-Visual Recognition of Emotional Engagement of People with Dementia}}, author = {Lars Steinert and Felix Putze and Dennis Küster and Tanja Schultz}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1024--1028}, doi = {10.21...
Dementia places an immeasurable burden on affected individuals and caregivers. In addition to general cognitive decline, dementia has a negative impact on communication. Technical activation systems are thus in high demand, as cognitive activation may help to moderate the decline. However, effective activation requires...
null
null
hecker21_interspeech
Speaking Corona? Human and Machine Recognition of COVID-19 from Voice
[ "Pascal Hecker", "Florian B. Pokorny", "Katrin D. Bartl-Pokorny", "Uwe Reichel", "Zhao Ren", "Simone Hantke", "Florian Eyben", "Dagmar M. Schuller", "Bert Arnrich", "Björn W. Schuller" ]
https://www.isca-archive.org/interspeech_2021/hecker21_interspeech.html
https://www.isca-archive.org/interspeech_2021/hecker21_interspeech.pdf
10.21437/Interspeech.2021-1771
1029-1033
@inproceedings{hecker21_interspeech, title = {{Speaking Corona? Human and Machine Recognition of COVID-19 from Voice}}, author = {Pascal Hecker and Florian B. Pokorny and Katrin D. Bartl-Pokorny and Uwe Reichel and Zhao Ren and Simone Hantke and Florian Eyben and Dagmar M. Schuller and Bert Arnrich and Björn...
With the COVID-19 pandemic, several research teams have reported successful advances in automated recognition of COVID-19 by voice. Resulting voice-based screening tools for COVID-19 could support large-scale testing efforts. While capabilities of machines on this task are progressing, we approach the so far unexplored...
null
null
nguyen21b_interspeech
Acoustic-Prosodic, Lexical and Demographic Cues to Persuasiveness in Competitive Debate Speeches
[ "Huyen Nguyen", "Ralph Vente", "David Lupea", "Sarah Ita Levitan", "Julia Hirschberg" ]
https://www.isca-archive.org/interspeech_2021/nguyen21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/nguyen21b_interspeech.pdf
10.21437/Interspeech.2021-1891
1034-1038
@inproceedings{nguyen21b_interspeech, title = {{Acoustic-Prosodic, Lexical and Demographic Cues to Persuasiveness in Competitive Debate Speeches}}, author = {Huyen Nguyen and Ralph Vente and David Lupea and Sarah Ita Levitan and Julia Hirschberg}, year = {2021}, booktitle = {{Interspeech 2021}}, p...
We analyze the acoustic-prosodic and lexical correlates of persuasiveness, taking into account speaker, judge and debate characteristics in a novel data set of 674 audio profiles, transcripts, evaluation scores and demographic data from professional debate tournament speeches. By conducting 10-fold cross validation exp...
null
null
borgstrom21_interspeech
Unsupervised Bayesian Adaptation of PLDA for Speaker Verification
[ "Bengt J. Borgström" ]
https://www.isca-archive.org/interspeech_2021/borgstrom21_interspeech.html
https://www.isca-archive.org/interspeech_2021/borgstrom21_interspeech.pdf
10.21437/Interspeech.2021-33
1039-1043
@inproceedings{borgstrom21_interspeech, title = {{Unsupervised Bayesian Adaptation of PLDA for Speaker Verification}}, author = {Bengt J. Borgström}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1039--1043}, doi = {10.21437/Interspeech.2021-33}, issn = {2958-1796}, ...
This paper presents a Bayesian framework for unsupervised domain adaptation of Probabilistic Linear Discriminant Analysis (PLDA). By interpreting class labels as latent random variables, Variational Bayes (VB) is used to derive a maximum a posterior (MAP) solution of the adapted PLDA model when labels are missing, refe...
null
null
wang21i_interspeech
The DKU-Duke-Lenovo System Description for the Fearless Steps Challenge Phase III
[ "Weiqing Wang", "Danwei Cai", "Jin Wang", "Qingjian Lin", "Xuyang Wang", "Mi Hong", "Ming Li" ]
https://www.isca-archive.org/interspeech_2021/wang21i_interspeech.html
https://www.isca-archive.org/interspeech_2021/wang21i_interspeech.pdf
10.21437/Interspeech.2021-235
1044-1048
@inproceedings{wang21i_interspeech, title = {{The DKU-Duke-Lenovo System Description for the Fearless Steps Challenge Phase III}}, author = {Weiqing Wang and Danwei Cai and Jin Wang and Qingjian Lin and Xuyang Wang and Mi Hong and Ming Li}, year = {2021}, booktitle = {{Interspeech 2021}}, pages ...
This paper describes the systems developed by the DKU-Duke-Lenovo team for the Fearless Steps Challenge Phase III. For the speech activity detection (SAD) task, we employ the U-Net-based model which has not been used for SAD before, observing a DCF of 1.915% on the eval set. For the speaker identification (SID) task, w...
null
null
chen21f_interspeech
Improved Meta-Learning Training for Speaker Verification
[ "Yafeng Chen", "Wu Guo", "Bin Gu" ]
https://www.isca-archive.org/interspeech_2021/chen21f_interspeech.html
https://www.isca-archive.org/interspeech_2021/chen21f_interspeech.pdf
10.21437/Interspeech.2021-405
1049-1053
@inproceedings{chen21f_interspeech, title = {{Improved Meta-Learning Training for Speaker Verification}}, author = {Yafeng Chen and Wu Guo and Bin Gu}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1049--1053}, doi = {10.21437/Interspeech.2021-405}, issn = {2958-1796...
Meta-learning (ML) has recently become a research hotspot in speaker verification (SV). We introduce two methods to improve the meta-learning training for SV in this paper. For the first method, a backbone embedding network is first jointly trained with the conventional cross entropy loss and prototypical networks (PN)...
2103.15421
title_snapshot
wang21j_interspeech
Variational Information Bottleneck Based Regularization for Speaker Recognition
[ "Dan Wang", "Yuanjie Dong", "Yaxing Li", "Yunfei Zi", "Zhihui Zhang", "Xiaoqi Li", "Shengwu Xiong" ]
https://www.isca-archive.org/interspeech_2021/wang21j_interspeech.html
https://www.isca-archive.org/interspeech_2021/wang21j_interspeech.pdf
10.21437/Interspeech.2021-482
1054-1058
@inproceedings{wang21j_interspeech, title = {{Variational Information Bottleneck Based Regularization for Speaker Recognition}}, author = {Dan Wang and Yuanjie Dong and Yaxing Li and Yunfei Zi and Zhihui Zhang and Xiaoqi Li and Shengwu Xiong}, year = {2021}, booktitle = {{Interspeech 2021}}, pages...
Speaker recognition (SR) is inevitably affected by noise in real-life scenarios, resulting in decreased recognition accuracy. In this paper, we introduce a novel regularization method, variable information bottleneck (VIB), in speaker recognition to extract robust speaker embeddings. VIB prompts the neural network to i...
null
null
brummer21_interspeech
Out of a Hundred Trials, How Many Errors Does Your Speaker Verifier Make?
[ "Niko Brümmer", "Luciana Ferrer", "Albert Swart" ]
https://www.isca-archive.org/interspeech_2021/brummer21_interspeech.html
https://www.isca-archive.org/interspeech_2021/brummer21_interspeech.pdf
10.21437/Interspeech.2021-541
1059-1063
@inproceedings{brummer21_interspeech, title = {{Out of a Hundred Trials, How Many Errors Does Your Speaker Verifier Make?}}, author = {Niko Brümmer and Luciana Ferrer and Albert Swart}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1059--1063}, doi = {10.21437/Interspeech.2...
Out of a hundred trials, how many errors does your speaker verifier make? For the user this is an important, practical question, but researchers and vendors typically sidestep it and supply instead the conditional error-rates that are given by the ROC/DET curve. We posit that the user’s question is answered by the Baye...
2104.00732
title_snapshot
chojnacka21_interspeech
SpeakerStew: Scaling to Many Languages with a Triaged Multilingual Text-Dependent and Text-Independent Speaker Verification System
[ "Roza Chojnacka", "Jason Pelecanos", "Quan Wang", "Ignacio Lopez Moreno" ]
https://www.isca-archive.org/interspeech_2021/chojnacka21_interspeech.html
https://www.isca-archive.org/interspeech_2021/chojnacka21_interspeech.pdf
10.21437/Interspeech.2021-646
1064-1068
@inproceedings{chojnacka21_interspeech, title = {{SpeakerStew: Scaling to Many Languages with a Triaged Multilingual Text-Dependent and Text-Independent Speaker Verification System}}, author = {Roza Chojnacka and Jason Pelecanos and Quan Wang and Ignacio Lopez Moreno}, year = {2021}, booktitle = {{I...
In this paper, we describe SpeakerStew — a hybrid system to perform speaker verification on 46 languages. Two core ideas were explored in this system: (1) Pooling training data of different languages together for multilingual generalization and reducing development cycles; (2) A novel triage mechanism between text-depe...
2104.02125
title_snapshot
wang21k_interspeech
AntVoice Neural Speaker Embedding System for FFSVC 2020
[ "Zhiming Wang", "Furong Xu", "Kaisheng Yao", "Yuan Cheng", "Tao Xiong", "Huijia Zhu" ]
https://www.isca-archive.org/interspeech_2021/wang21k_interspeech.html
https://www.isca-archive.org/interspeech_2021/wang21k_interspeech.pdf
10.21437/Interspeech.2021-966
1069-1073
@inproceedings{wang21k_interspeech, title = {{AntVoice Neural Speaker Embedding System for FFSVC 2020}}, author = {Zhiming Wang and Furong Xu and Kaisheng Yao and Yuan Cheng and Tao Xiong and Huijia Zhu}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1069--1073}, doi = {10....
This paper presents a comprehensive description of the AntVoice system for the first two tracks of far-field speaker verification from single microphone array in FFSVC 2020 [1]. The system is based on neural speaker embeddings from deep neural network-based encoder networks. These encoder networks for acoustic modeling...
null
null
li21b_interspeech
Gradient Regularization for Noise-Robust Speaker Verification
[ "Jianchen Li", "Jiqing Han", "Hongwei Song" ]
https://www.isca-archive.org/interspeech_2021/li21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/li21b_interspeech.pdf
10.21437/Interspeech.2021-1216
1074-1078
@inproceedings{li21b_interspeech, title = {{Gradient Regularization for Noise-Robust Speaker Verification}}, author = {Jianchen Li and Jiqing Han and Hongwei Song}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1074--1078}, doi = {10.21437/Interspeech.2021-1216}, issn ...
Noise robustness is a challenge for speaker recognition systems. To solve this problem, one of the most common approaches is to joint-train a model by using both clean and noisy utterances. However, the gradients calculated on noisy utterances generally contain speaker-irrelevant noisy components, resulting in overfitt...
null
null
kataria21_interspeech
Deep Feature CycleGANs: Speaker Identity Preserving Non-Parallel Microphone-Telephone Domain Adaptation for Speaker Verification
[ "Saurabh Kataria", "Jesús Villalba", "Piotr Żelasko", "Laureano Moro-Velázquez", "Najim Dehak" ]
https://www.isca-archive.org/interspeech_2021/kataria21_interspeech.html
https://www.isca-archive.org/interspeech_2021/kataria21_interspeech.pdf
10.21437/Interspeech.2021-1502
1079-1083
@inproceedings{kataria21_interspeech, title = {{Deep Feature CycleGANs: Speaker Identity Preserving Non-Parallel Microphone-Telephone Domain Adaptation for Speaker Verification}}, author = {Saurabh Kataria and Jesús Villalba and Piotr Żelasko and Laureano Moro-Velázquez and Najim Dehak}, year = {2021}...
With the increase in the availability of speech from varied domains, it is imperative to use such out-of-domain data to improve existing speech systems. Domain adaptation is a prominent pre-processing approach for this. We investigate it to adapt microphone speech to the telephone domain. Specifically, we explore Cycle...
2104.01433
title_snapshot
pu21_interspeech
Scaling Effect of Self-Supervised Speech Models
[ "Jie Pu", "Yuguang Yang", "Ruirui Li", "Oguz Elibol", "Jasha Droppo" ]
https://www.isca-archive.org/interspeech_2021/pu21_interspeech.html
https://www.isca-archive.org/interspeech_2021/pu21_interspeech.pdf
10.21437/Interspeech.2021-1935
1084-1088
@inproceedings{pu21_interspeech, title = {{Scaling Effect of Self-Supervised Speech Models}}, author = {Jie Pu and Yuguang Yang and Ruirui Li and Oguz Elibol and Jasha Droppo}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1084--1088}, doi = {10.21437/Interspeech.2021-1935}...
The success of modern deep learning systems is built on two cornerstones, massive amount of annotated training data and advanced computational infrastructure to support large-scale computation. In recent years, the model size of state-of-the-art deep learning systems has rapidly increased and sometimes reached to billi...
null
null
wu21c_interspeech
Joint Feature Enhancement and Speaker Recognition with Multi-Objective Task-Oriented Network
[ "Yibo Wu", "Longbiao Wang", "Kong Aik Lee", "Meng Liu", "Jianwu Dang" ]
https://www.isca-archive.org/interspeech_2021/wu21c_interspeech.html
https://www.isca-archive.org/interspeech_2021/wu21c_interspeech.pdf
10.21437/Interspeech.2021-1978
1089-1093
@inproceedings{wu21c_interspeech, title = {{Joint Feature Enhancement and Speaker Recognition with Multi-Objective Task-Oriented Network}}, author = {Yibo Wu and Longbiao Wang and Kong Aik Lee and Meng Liu and Jianwu Dang}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1089--1093},...
Recently, increasing attention has been paid to the joint training of upstream and downstream tasks, and to address the challenge of how to synchronize various loss functions in a multi-objective scenario. In this paper, to address the competing gradient directions between the speaker classification loss and the featur...
null
null
zhang21g_interspeech
Multi-Level Transfer Learning from Near-Field to Far-Field Speaker Verification
[ "Li Zhang", "Qing Wang", "Kong Aik Lee", "Lei Xie", "Haizhou Li" ]
https://www.isca-archive.org/interspeech_2021/zhang21g_interspeech.html
https://www.isca-archive.org/interspeech_2021/zhang21g_interspeech.pdf
10.21437/Interspeech.2021-1980
1094-1098
@inproceedings{zhang21g_interspeech, title = {{Multi-Level Transfer Learning from Near-Field to Far-Field Speaker Verification}}, author = {Li Zhang and Qing Wang and Kong Aik Lee and Lei Xie and Haizhou Li}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1094--1098}, doi = ...
In far-field speaker verification, the performance of speaker embeddings is susceptible to degradation when there is a mismatch between the conditions of enrollment and test speech. To solve this problem, we propose the feature-level and instance-level transfer learning in the teacher-student framework to learn a domai...
2106.09320
title_snapshot
patino21_interspeech
Speaker Anonymisation Using the McAdams Coefficient
[ "Jose Patino", "Natalia Tomashenko", "Massimiliano Todisco", "Andreas Nautsch", "Nicholas Evans" ]
https://www.isca-archive.org/interspeech_2021/patino21_interspeech.html
https://www.isca-archive.org/interspeech_2021/patino21_interspeech.pdf
10.21437/Interspeech.2021-1070
1099-1103
@inproceedings{patino21_interspeech, title = {{Speaker Anonymisation Using the McAdams Coefficient}}, author = {Jose Patino and Natalia Tomashenko and Massimiliano Todisco and Andreas Nautsch and Nicholas Evans}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1099--1103}, doi ...
Anonymisation has the goal of manipulating speech signals in order to degrade the reliability of automatic approaches to speaker recognition, while preserving other aspects of speech, such as those relating to intelligibility and naturalness. This paper reports an approach to anonymisation that, unlike other current ap...
2011.01130
title_snapshot
luo21_interspeech
Multi-Stream Gated and Pyramidal Temporal Convolutional Neural Networks for Audio-Visual Speech Separation in Multi-Talker Environments
[ "Yiyu Luo", "Jing Wang", "Liang Xu", "Lidong Yang" ]
https://www.isca-archive.org/interspeech_2021/luo21_interspeech.html
https://www.isca-archive.org/interspeech_2021/luo21_interspeech.pdf
10.21437/Interspeech.2021-366
1104-1108
@inproceedings{luo21_interspeech, title = {{Multi-Stream Gated and Pyramidal Temporal Convolutional Neural Networks for Audio-Visual Speech Separation in Multi-Talker Environments}}, author = {Yiyu Luo and Jing Wang and Liang Xu and Lidong Yang}, year = {2021}, booktitle = {{Interspeech 2021}}, pa...
Speech separation is the task of extracting target speech from noisy mixture. In applications like video telephones or video conferencing, lip movements of the target speaker are accessible, which can be leveraged for speech separation. This paper proposes a time-domain audio-visual speech separation model under multi-...
null
null
wang21l_interspeech
TeCANet: Temporal-Contextual Attention Network for Environment-Aware Speech Dereverberation
[ "Helin Wang", "Bo Wu", "Lianwu Chen", "Meng Yu", "Jianwei Yu", "Yong Xu", "Shi-Xiong Zhang", "Chao Weng", "Dan Su", "Dong Yu" ]
https://www.isca-archive.org/interspeech_2021/wang21l_interspeech.html
https://www.isca-archive.org/interspeech_2021/wang21l_interspeech.pdf
10.21437/Interspeech.2021-481
1109-1113
@inproceedings{wang21l_interspeech, title = {{TeCANet: Temporal-Contextual Attention Network for Environment-Aware Speech Dereverberation}}, author = {Helin Wang and Bo Wu and Lianwu Chen and Meng Yu and Jianwei Yu and Yong Xu and Shi-Xiong Zhang and Chao Weng and Dan Su and Dong Yu}, year = {2021}, ...
In this paper, we exploit the effective way to leverage contextual information to improve the speech dereverberation performance in real-world reverberant environments. We propose a temporal-contextual attention approach on the deep neural network (DNN) for environment-aware speech dereverberation, which can adaptively...
2103.16849
title_snapshot
gu21_interspeech
Residual Echo and Noise Cancellation with Feature Attention Module and Multi-Domain Loss Function
[ "Jianjun Gu", "Longbiao Cheng", "Xingwei Sun", "Junfeng Li", "Yonghong Yan" ]
https://www.isca-archive.org/interspeech_2021/gu21_interspeech.html
https://www.isca-archive.org/interspeech_2021/gu21_interspeech.pdf
10.21437/Interspeech.2021-538
1114-1118
@inproceedings{gu21_interspeech, title = {{Residual Echo and Noise Cancellation with Feature Attention Module and Multi-Domain Loss Function}}, author = {Jianjun Gu and Longbiao Cheng and Xingwei Sun and Junfeng Li and Yonghong Yan}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {11...
For real-time acoustic echo cancellation in noisy environments, the classical linear adaptive filters (LAFs) can only remove the linear components of acoustic echo. To further attenuate the non-linear echo components and background noise, this paper proposes a deep learning-based residual echo and noise cancellation (R...
null
null
li21c_interspeech
MIMO Self-Attentive RNN Beamformer for Multi-Speaker Speech Separation
[ "Xiyun Li", "Yong Xu", "Meng Yu", "Shi-Xiong Zhang", "Jiaming Xu", "Bo Xu", "Dong Yu" ]
https://www.isca-archive.org/interspeech_2021/li21c_interspeech.html
https://www.isca-archive.org/interspeech_2021/li21c_interspeech.pdf
10.21437/Interspeech.2021-570
1119-1123
@inproceedings{li21c_interspeech, title = {{MIMO Self-Attentive RNN Beamformer for Multi-Speaker Speech Separation}}, author = {Xiyun Li and Yong Xu and Meng Yu and Shi-Xiong Zhang and Jiaming Xu and Bo Xu and Dong Yu}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1119--1123}, d...
Recently, our proposed recurrent neural network (RNN) based all deep learning minimum variance distortionless response (ADL-MVDR) beamformer method yielded superior performance over the conventional MVDR by replacing the matrix inversion and eigenvalue decomposition with two RNNs. In this work, we present a self-attent...
2104.08450
title_snapshot
giri21_interspeech
Personalized PercepNet: Real-Time, Low-Complexity Target Voice Separation and Enhancement
[ "Ritwik Giri", "Shrikant Venkataramani", "Jean-Marc Valin", "Umut Isik", "Arvindh Krishnaswamy" ]
https://www.isca-archive.org/interspeech_2021/giri21_interspeech.html
https://www.isca-archive.org/interspeech_2021/giri21_interspeech.pdf
10.21437/Interspeech.2021-694
1124-1128
@inproceedings{giri21_interspeech, title = {{Personalized PercepNet: Real-Time, Low-Complexity Target Voice Separation and Enhancement}}, author = {Ritwik Giri and Shrikant Venkataramani and Jean-Marc Valin and Umut Isik and Arvindh Krishnaswamy}, year = {2021}, booktitle = {{Interspeech 2021}}, p...
The presence of multiple talkers in the surrounding environment poses a difficult challenge for real-time speech communication systems considering the constraints on network size and complexity. In this paper, we present Personalized PercepNet, a real-time speech enhancement model that separates a target speaker from a...
2106.04129
title_snapshot
yemini21_interspeech
Scene-Agnostic Multi-Microphone Speech Dereverberation
[ "Yochai Yemini", "Ethan Fetaya", "Haggai Maron", "Sharon Gannot" ]
https://www.isca-archive.org/interspeech_2021/yemini21_interspeech.html
https://www.isca-archive.org/interspeech_2021/yemini21_interspeech.pdf
10.21437/Interspeech.2021-889
1129-1133
@inproceedings{yemini21_interspeech, title = {{Scene-Agnostic Multi-Microphone Speech Dereverberation}}, author = {Yochai Yemini and Ethan Fetaya and Haggai Maron and Sharon Gannot}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1129--1133}, doi = {10.21437/Interspeech.2021...
Neural networks (NNs) have been widely applied in speech processing tasks, and, in particular, those employing microphone arrays. Nevertheless, most existing NN architectures can only deal with fixed and position-specific microphone arrays. In this paper, we present an NN architecture that can cope with microphone arra...
2010.11875
title_snapshot
tanaka21_interspeech
Manifold-Aware Deep Clustering: Maximizing Angles Between Embedding Vectors Based on Regular Simplex
[ "Keitaro Tanaka", "Ryosuke Sawata", "Shusuke Takahashi" ]
https://www.isca-archive.org/interspeech_2021/tanaka21_interspeech.html
https://www.isca-archive.org/interspeech_2021/tanaka21_interspeech.pdf
10.21437/Interspeech.2021-1029
1134-1138
@inproceedings{tanaka21_interspeech, title = {{Manifold-Aware Deep Clustering: Maximizing Angles Between Embedding Vectors Based on Regular Simplex}}, author = {Keitaro Tanaka and Ryosuke Sawata and Shusuke Takahashi}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1134--1138}, do...
This paper presents a new deep clustering (DC) method called manifold-aware DC (M-DC) that can enhance hyperspace utilization more effectively than the original DC. The original DC has a limitation in that a pair of two speakers has to be embedded having an orthogonal relationship due to its use of the one-hot vector-b...
2106.02331
title_snapshot
zhang21h_interspeech
A Deep Learning Approach to Multi-Channel and Multi-Microphone Acoustic Echo Cancellation
[ "Hao Zhang", "DeLiang Wang" ]
https://www.isca-archive.org/interspeech_2021/zhang21h_interspeech.html
https://www.isca-archive.org/interspeech_2021/zhang21h_interspeech.pdf
10.21437/Interspeech.2021-1508
1139-1143
@inproceedings{zhang21h_interspeech, title = {{A Deep Learning Approach to Multi-Channel and Multi-Microphone Acoustic Echo Cancellation}}, author = {Hao Zhang and DeLiang Wang}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1139--1143}, doi = {10.21437/Interspeech.2021-150...
Building on deep learning based acoustic echo cancellation (AEC) in the single-loudspeaker (single-channel) and single-microphone setup, this paper investigates multi-channel (multi-loudspeaker) AEC (MCAEC) and multi-microphone AEC (MMAEC). A convolutional recurrent network (CRN) is trained to predict the near-end spee...
2103.02552
title_judge
na21_interspeech
Joint Online Multichannel Acoustic Echo Cancellation, Speech Dereverberation and Source Separation
[ "Yueyue Na", "Ziteng Wang", "Zhang Liu", "Biao Tian", "Qiang Fu" ]
https://www.isca-archive.org/interspeech_2021/na21_interspeech.html
https://www.isca-archive.org/interspeech_2021/na21_interspeech.pdf
10.21437/Interspeech.2021-1950
1144-1148
@inproceedings{na21_interspeech, title = {{Joint Online Multichannel Acoustic Echo Cancellation, Speech Dereverberation and Source Separation}}, author = {Yueyue Na and Ziteng Wang and Zhang Liu and Biao Tian and Qiang Fu}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1144--1148},...
This paper presents a joint source separation algorithm that simultaneously reduces acoustic echo, reverberation and interfering sources. Target speeches are separated from the mixture by maximizing independence with respect to the other sources. It is shown that the separation process can be decomposed into cascading ...
2104.04325
title_snapshot
sato21_interspeech
Should We Always Separate?: Switching Between Enhanced and Observed Signals for Overlapping Speech Recognition
[ "Hiroshi Sato", "Tsubasa Ochiai", "Marc Delcroix", "Keisuke Kinoshita", "Takafumi Moriya", "Naoyuki Kamo" ]
https://www.isca-archive.org/interspeech_2021/sato21_interspeech.html
https://www.isca-archive.org/interspeech_2021/sato21_interspeech.pdf
10.21437/Interspeech.2021-2253
1149-1153
@inproceedings{sato21_interspeech, title = {{Should We Always Separate?: Switching Between Enhanced and Observed Signals for Overlapping Speech Recognition}}, author = {Hiroshi Sato and Tsubasa Ochiai and Marc Delcroix and Keisuke Kinoshita and Takafumi Moriya and Naoyuki Kamo}, year = {2021}, bookt...
Although recent advances in deep learning technology improved automatic speech recognition (ASR), it remains difficult to recognize speech when it overlaps other people’s voices. Speech separation or extraction is often used as a front-end to ASR to handle such overlapping speech. However, deep neural network-based spe...
2106.00949
title_snapshot
udupa21_interspeech
Estimating Articulatory Movements in Speech Production with Transformer Networks
[ "Sathvik Udupa", "Anwesha Roy", "Abhayjeet Singh", "Aravind Illa", "Prasanta Kumar Ghosh" ]
https://www.isca-archive.org/interspeech_2021/udupa21_interspeech.html
https://www.isca-archive.org/interspeech_2021/udupa21_interspeech.pdf
10.21437/Interspeech.2021-1375
1154-1158
@inproceedings{udupa21_interspeech, title = {{Estimating Articulatory Movements in Speech Production with Transformer Networks}}, author = {Sathvik Udupa and Anwesha Roy and Abhayjeet Singh and Aravind Illa and Prasanta Kumar Ghosh}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {11...
We estimate articulatory movements in speech production from different modalities - acoustics and phonemes. Acoustic-to-articulatory inversion (AAI) is a sequence-to-sequence task. On the other hand, phoneme to articulatory (PTA) motion estimation faces a key challenge in reliably aligning the text and the articulatory...
2104.05017
title_snapshot
yang21b_interspeech
Unsupervised Multi-Target Domain Adaptation for Acoustic Scene Classification
[ "Dongchao Yang", "Helin Wang", "Yuexian Zou" ]
https://www.isca-archive.org/interspeech_2021/yang21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/yang21b_interspeech.pdf
10.21437/Interspeech.2021-300
1159-1163
@inproceedings{yang21b_interspeech, title = {{Unsupervised Multi-Target Domain Adaptation for Acoustic Scene Classification}}, author = {Dongchao Yang and Helin Wang and Yuexian Zou}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1159--1163}, doi = {10.21437/Interspeech.202...
It is well known that the mismatch between training (source) and test (target) data distribution will significantly decrease the performance of acoustic scene classification (ASC) systems. To address this issue, domain adaptation (DA) is one solution and many unsupervised DA methods have been proposed. These methods fo...
2105.10340
title_snapshot
jaramillo21_interspeech
Speech Decomposition Based on a Hybrid Speech Model and Optimal Segmentation
[ "Alfredo Esquivel Jaramillo", "Jesper Kjær Nielsen", "Mads Græsbøll Christensen" ]
https://www.isca-archive.org/interspeech_2021/jaramillo21_interspeech.html
https://www.isca-archive.org/interspeech_2021/jaramillo21_interspeech.pdf
10.21437/Interspeech.2021-47
1164-1168
@inproceedings{jaramillo21_interspeech, title = {{Speech Decomposition Based on a Hybrid Speech Model and Optimal Segmentation}}, author = {Alfredo Esquivel Jaramillo and Jesper Kjær Nielsen and Mads Græsbøll Christensen}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1164--1168}, ...
In a hybrid speech model, both voiced and unvoiced components can coexist in a segment. Often, the voiced speech is regarded as the deterministic component, and the unvoiced speech and additive noise are the stochastic components. Typically, the speech signal is considered stationary within fixed segments of 20–40 ms, ...
2105.01302
title_snapshot
luo21b_interspeech
Dropout Regularization for Self-Supervised Learning of Transformer Encoder Speech Representation
[ "Jian Luo", "Jianzong Wang", "Ning Cheng", "Jing Xiao" ]
https://www.isca-archive.org/interspeech_2021/luo21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/luo21b_interspeech.pdf
10.21437/Interspeech.2021-1066
1169-1173
@inproceedings{luo21b_interspeech, title = {{Dropout Regularization for Self-Supervised Learning of Transformer Encoder Speech Representation}}, author = {Jian Luo and Jianzong Wang and Ning Cheng and Jing Xiao}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1169--1173}, doi ...
Predicting the altered acoustic frames is an effective way of self-supervised learning for speech representation. However, it is challenging to prevent the pretrained model from overfitting. In this paper, we proposed to introduce two dropout regularization methods into the pretraining of transformer encoder: (1) atten...
2107.04227
title_snapshot
yarra21_interspeech
Noise Robust Pitch Stylization Using Minimum Mean Absolute Error Criterion
[ "Chiranjeevi Yarra", "Prasanta Kumar Ghosh" ]
https://www.isca-archive.org/interspeech_2021/yarra21_interspeech.html
https://www.isca-archive.org/interspeech_2021/yarra21_interspeech.pdf
10.21437/Interspeech.2021-1307
1174-1178
@inproceedings{yarra21_interspeech, title = {{Noise Robust Pitch Stylization Using Minimum Mean Absolute Error Criterion}}, author = {Chiranjeevi Yarra and Prasanta Kumar Ghosh}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1174--1178}, doi = {10.21437/Interspeech.2021-130...
We propose a pitch stylization technique in the presence of pitch halving and doubling errors. The technique uses an optimization criterion based on a minimum mean absolute error to make the stylization robust to such pitch estimation errors, particularly under noisy conditions. We obtain segments for the stylization a...
null
null
huang21b_interspeech
An Attribute-Aligned Strategy for Learning Speech Representation
[ "Yu-Lin Huang", "Bo-Hao Su", "Y.-W. Peter Hong", "Chi-Chun Lee" ]
https://www.isca-archive.org/interspeech_2021/huang21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/huang21b_interspeech.pdf
10.21437/Interspeech.2021-1341
1179-1183
@inproceedings{huang21b_interspeech, title = {{An Attribute-Aligned Strategy for Learning Speech Representation}}, author = {Yu-Lin Huang and Bo-Hao Su and Y.-W. Peter Hong and Chi-Chun Lee}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1179--1183}, doi = {10.21437/Intersp...
Advancement in speech technology has brought convenience to our life. However, the concern is on the rise as speech signal contains multiple personal attributes, which would lead to either sensitive information leakage or bias toward decision. In this work, we propose an attribute-aligned learning strategy to derive sp...
2106.02810
title_snapshot
shahrebabaki21_interspeech
Raw Speech-to-Articulatory Inversion by Temporal Filtering and Decimation
[ "Abdolreza Sabzi Shahrebabaki", "Sabato Marco Siniscalchi", "Torbjørn Svendsen" ]
https://www.isca-archive.org/interspeech_2021/shahrebabaki21_interspeech.html
https://www.isca-archive.org/interspeech_2021/shahrebabaki21_interspeech.pdf
10.21437/Interspeech.2021-1429
1184-1188
@inproceedings{shahrebabaki21_interspeech, title = {{Raw Speech-to-Articulatory Inversion by Temporal Filtering and Decimation}}, author = {Abdolreza Sabzi Shahrebabaki and Sabato Marco Siniscalchi and Torbjørn Svendsen}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1184--1188}, ...
We propose a novel sequence-to-sequence acoustic-to-articulatory inversion (AAI) neural architecture in the temporal waveform domain. In contrast to traditional AAI approaches that leverage hand-crafted short-time spectral features obtained from the windowed signal, such as LSFs, or MFCCs, our solution directly process...
null
null
lilley21_interspeech
Unsupervised Training of a DNN-Based Formant Tracker
[ "Jason Lilley", "H. Timothy Bunnell" ]
https://www.isca-archive.org/interspeech_2021/lilley21_interspeech.html
https://www.isca-archive.org/interspeech_2021/lilley21_interspeech.pdf
10.21437/Interspeech.2021-1690
1189-1193
@inproceedings{lilley21_interspeech, title = {{Unsupervised Training of a DNN-Based Formant Tracker}}, author = {Jason Lilley and H. Timothy Bunnell}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1189--1193}, doi = {10.21437/Interspeech.2021-1690}, issn = {2958-1796...
Phonetic analysis often requires reliable estimation of formants, but estimates provided by popular programs can be unreliable. Recently, Dissen et al. [1] described DNN-based formant trackers that produced more accurate frequency estimates than several others, but require manually-corrected formant data for training. ...
null
null
yang21c_interspeech
SUPERB: Speech Processing Universal PERformance Benchmark
[ "Shu-wen Yang", "Po-Han Chi", "Yung-Sung Chuang", "Cheng-I Jeff Lai", "Kushal Lakhotia", "Yist Y. Lin", "Andy T. Liu", "Jiatong Shi", "Xuankai Chang", "Guan-Ting Lin", "Tzu-Hsien Huang", "Wei-Cheng Tseng", "Ko-tik Lee", "Da-Rong Liu", "Zili Huang", "Shuyan Dong", "Shang-Wen Li", "S...
https://www.isca-archive.org/interspeech_2021/yang21c_interspeech.html
https://www.isca-archive.org/interspeech_2021/yang21c_interspeech.pdf
10.21437/Interspeech.2021-1775
1194-1198
@inproceedings{yang21c_interspeech, title = {{SUPERB: Speech Processing Universal PERformance Benchmark}}, author = {Shu-wen Yang and Po-Han Chi and Yung-Sung Chuang and Cheng-I Jeff Lai and Kushal Lakhotia and Yist Y. Lin and Andy T. Liu and Jiatong Shi and Xuankai Chang and Guan-Ting Lin and Tzu-Hsien Huan...
Self-supervised learning (SSL) has proven vital for advancing research in natural language processing (NLP) and computer vision (CV). The paradigm pretrains a shared model on large volumes of unlabeled data and achieves state-of-the-art (SOTA) for various tasks with minimal adaptation . However, the speech processing c...
2105.01051
title_snapshot
zhang21i_interspeech
Synchronising Speech Segments with Musical Beats in Mandarin and English Singing
[ "Cong Zhang", "Jian Zhu" ]
https://www.isca-archive.org/interspeech_2021/zhang21i_interspeech.html
https://www.isca-archive.org/interspeech_2021/zhang21i_interspeech.pdf
10.21437/Interspeech.2021-1841
1199-1203
@inproceedings{zhang21i_interspeech, title = {{Synchronising Speech Segments with Musical Beats in Mandarin and English Singing}}, author = {Cong Zhang and Jian Zhu}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1199--1203}, doi = {10.21437/Interspeech.2021-1841}, issn ...
Generating synthesised singing voice with models trained on speech data has many advantages due to the models’ flexibility and controllability. However, since the information about the temporal relationship between segments and beats are lacking in speech training data, the synthesised singing may sound off-beat at tim...
2106.10045
title_snapshot
peplinski21_interspeech
FRILL: A Non-Semantic Speech Embedding for Mobile Devices
[ "Jacob Peplinski", "Joel Shor", "Sachin Joglekar", "Jake Garrison", "Shwetak Patel" ]
https://www.isca-archive.org/interspeech_2021/peplinski21_interspeech.html
https://www.isca-archive.org/interspeech_2021/peplinski21_interspeech.pdf
10.21437/Interspeech.2021-2070
1204-1208
@inproceedings{peplinski21_interspeech, title = {{FRILL: A Non-Semantic Speech Embedding for Mobile Devices}}, author = {Jacob Peplinski and Joel Shor and Sachin Joglekar and Jake Garrison and Shwetak Patel}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1204--1208}, doi = ...
Learned speech representations can drastically improve performance on tasks with limited labeled data. However, due to their size and complexity, learned representations have limited utility in mobile settings where run-time performance can be a significant bottleneck. In this work, we propose a class of lightweight no...
2011.04609
title_snapshot
mori21_interspeech
Pitch Contour Separation from Overlapping Speech
[ "Hiroki Mori" ]
https://www.isca-archive.org/interspeech_2021/mori21_interspeech.html
https://www.isca-archive.org/interspeech_2021/mori21_interspeech.pdf
10.21437/Interspeech.2021-2164
1209-1213
@inproceedings{mori21_interspeech, title = {{Pitch Contour Separation from Overlapping Speech}}, author = {Hiroki Mori}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1209--1213}, doi = {10.21437/Interspeech.2021-2164}, issn = {2958-1796}, }
In everyday conversation, speakers’ utterances often overlap. For conversation corpora that are recorded in diverse environments, results of pitch extraction in the overlapping parts may be incorrect. The goal of this study is to establish the technique of separating each speaker’s pitch contour from an overlapping spe...
null
null
kumar21_interspeech
Do Sound Event Representations Generalize to Other Audio Tasks? A Case Study in Audio Transfer Learning
[ "Anurag Kumar", "Yun Wang", "Vamsi Krishna Ithapu", "Christian Fuegen" ]
https://www.isca-archive.org/interspeech_2021/kumar21_interspeech.html
https://www.isca-archive.org/interspeech_2021/kumar21_interspeech.pdf
10.21437/Interspeech.2021-347
1214-1218
@inproceedings{kumar21_interspeech, title = {{Do Sound Event Representations Generalize to Other Audio Tasks? A Case Study in Audio Transfer Learning}}, author = {Anurag Kumar and Yun Wang and Vamsi Krishna Ithapu and Christian Fuegen}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = ...
Transfer learning is critical for efficient information transfer across multiple related learning problems. A simple, yet effective transfer learning approach utilizes deep neural networks trained on a large-scale task for feature extraction. Such representations are then used to learn related downstream tasks. In this...
2106.11335
title_snapshot
peng21b_interspeech
Data Augmentation for Spoken Language Understanding via Pretrained Language Models
[ "Baolin Peng", "Chenguang Zhu", "Michael Zeng", "Jianfeng Gao" ]
https://www.isca-archive.org/interspeech_2021/peng21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/peng21b_interspeech.pdf
10.21437/Interspeech.2021-117
1219-1223
@inproceedings{peng21b_interspeech, title = {{Data Augmentation for Spoken Language Understanding via Pretrained Language Models}}, author = {Baolin Peng and Chenguang Zhu and Michael Zeng and Jianfeng Gao}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1219--1223}, doi = {...
The training of spoken language understanding (SLU) models often faces the problem of data scarcity. In this paper, we put forward a data augmentation method using pretrained language models to boost the variability and accuracy of generated utterances. Furthermore, we investigate and propose solutions to two previousl...
2004.13952
title_snapshot
radfar21_interspeech
FANS: Fusing ASR and NLU for On-Device SLU
[ "Martin Radfar", "Athanasios Mouchtaris", "Siegfried Kunzmann", "Ariya Rastrow" ]
https://www.isca-archive.org/interspeech_2021/radfar21_interspeech.html
https://www.isca-archive.org/interspeech_2021/radfar21_interspeech.pdf
10.21437/Interspeech.2021-793
1224-1228
@inproceedings{radfar21_interspeech, title = {{FANS: Fusing ASR and NLU for On-Device SLU}}, author = {Martin Radfar and Athanasios Mouchtaris and Siegfried Kunzmann and Ariya Rastrow}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1224--1228}, doi = {10.21437/Interspeech.2...
Spoken language understanding (SLU) systems translate voice input commands to semantics which are encoded as an intent and pairs of slot tags and values. Most current SLU systems deploy a cascade of two neural models where the first one maps the input audio to a transcript (ASR) and the second predicts the intent and s...
2111.00400
title_snapshot
cao21c_interspeech
Sequential End-to-End Intent and Slot Label Classification and Localization
[ "Yiran Cao", "Nihal Potdar", "Anderson R. Avila" ]
https://www.isca-archive.org/interspeech_2021/cao21c_interspeech.html
https://www.isca-archive.org/interspeech_2021/cao21c_interspeech.pdf
10.21437/Interspeech.2021-1569
1229-1233
@inproceedings{cao21c_interspeech, title = {{Sequential End-to-End Intent and Slot Label Classification and Localization}}, author = {Yiran Cao and Nihal Potdar and Anderson R. Avila}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1229--1233}, doi = {10.21437/Interspeech.20...
Human-computer interaction (HCI) is significantly impacted by delayed responses from a spoken dialogue system. Hence, end-to-end (e2e) spoken language understanding (SLU) solutions have recently been proposed to decrease latency. Such approaches allow for the extraction of semantic information directly from the speech ...
2106.04660
title_snapshot
muralidharan21_interspeech
DEXTER: Deep Encoding of External Knowledge for Named Entity Recognition in Virtual Assistants
[ "Deepak Muralidharan", "Joel Ruben Antony Moniz", "Weicheng Zhang", "Stephen Pulman", "Lin Li", "Megan Barnes", "Jingjing Pan", "Jason Williams", "Alex Acero" ]
https://www.isca-archive.org/interspeech_2021/muralidharan21_interspeech.html
https://www.isca-archive.org/interspeech_2021/muralidharan21_interspeech.pdf
10.21437/Interspeech.2021-1877
1234-1238
@inproceedings{muralidharan21_interspeech, title = {{DEXTER: Deep Encoding of External Knowledge for Named Entity Recognition in Virtual Assistants}}, author = {Deepak Muralidharan and Joel Ruben Antony Moniz and Weicheng Zhang and Stephen Pulman and Lin Li and Megan Barnes and Jingjing Pan and Jason William...
Named entity recognition (NER) is usually developed and tested on text from well-written sources. However, in intelligent voice assistants, where NER is an important component, input to NER may be noisy because of user or speech recognition error. In applications, entity labels may change frequently, and non-textual pr...
2108.06633
title_snapshot
wu21d_interspeech
A Context-Aware Hierarchical BERT Fusion Network for Multi-Turn Dialog Act Detection
[ "Ting-Wei Wu", "Ruolin Su", "Biing-Hwang Juang" ]
https://www.isca-archive.org/interspeech_2021/wu21d_interspeech.html
https://www.isca-archive.org/interspeech_2021/wu21d_interspeech.pdf
10.21437/Interspeech.2021-95
1239-1243
@inproceedings{wu21d_interspeech, title = {{A Context-Aware Hierarchical BERT Fusion Network for Multi-Turn Dialog Act Detection}}, author = {Ting-Wei Wu and Ruolin Su and Biing-Hwang Juang}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1239--1243}, doi = {10.21437/Intersp...
The success of interactive dialog systems is usually associated with the quality of the spoken language understanding (SLU) task, which mainly identifies the corresponding dialog acts and slot values in each turn. By treating utterances in isolation, most SLU systems often overlook the semantic context in which a dialo...
2109.01267
title_snapshot
chen21g_interspeech
Pre-Training for Spoken Language Understanding with Joint Textual and Phonetic Representation Learning
[ "Qian Chen", "Wen Wang", "Qinglin Zhang" ]
https://www.isca-archive.org/interspeech_2021/chen21g_interspeech.html
https://www.isca-archive.org/interspeech_2021/chen21g_interspeech.pdf
10.21437/Interspeech.2021-234
1244-1248
@inproceedings{chen21g_interspeech, title = {{Pre-Training for Spoken Language Understanding with Joint Textual and Phonetic Representation Learning}}, author = {Qian Chen and Wen Wang and Qinglin Zhang}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1244--1248}, doi = {10....
In the traditional cascading architecture for spoken language understanding (SLU), it has been observed that automatic speech recognition errors could be detrimental to the performance of natural language understanding. End-to-end (E2E) SLU models have been proposed to directly map speech input to desired semantic fram...
2104.10357
title_snapshot
do21b_interspeech
Predicting Temporal Performance Drop of Deployed Production Spoken Language Understanding Models
[ "Quynh Do", "Judith Gaspers", "Daniil Sorokin", "Patrick Lehnen" ]
https://www.isca-archive.org/interspeech_2021/do21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/do21b_interspeech.pdf
10.21437/Interspeech.2021-580
1249-1253
@inproceedings{do21b_interspeech, title = {{Predicting Temporal Performance Drop of Deployed Production Spoken Language Understanding Models}}, author = {Quynh Do and Judith Gaspers and Daniil Sorokin and Patrick Lehnen}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1249--1253}, ...
In deployed real-world spoken language understanding (SLU) applications, data continuously flows into the system. This leads to distributional differences between training and application data that can deteriorate model performance. While regularly retraining the deployed model with new data helps mitigating this probl...
null
null
ganhotra21_interspeech
Integrating Dialog History into End-to-End Spoken Language Understanding Systems
[ "Jatin Ganhotra", "Samuel Thomas", "Hong-Kwang J. Kuo", "Sachindra Joshi", "George Saon", "Zoltán Tüske", "Brian Kingsbury" ]
https://www.isca-archive.org/interspeech_2021/ganhotra21_interspeech.html
https://www.isca-archive.org/interspeech_2021/ganhotra21_interspeech.pdf
10.21437/Interspeech.2021-1460
1254-1258
@inproceedings{ganhotra21_interspeech, title = {{Integrating Dialog History into End-to-End Spoken Language Understanding Systems}}, author = {Jatin Ganhotra and Samuel Thomas and Hong-Kwang J. Kuo and Sachindra Joshi and George Saon and Zoltán Tüske and Brian Kingsbury}, year = {2021}, booktitle = ...
End-to-end spoken language understanding (SLU) systems that process human-human or human-computer interactions are often context independent and process each turn of a conversation independently. Spoken conversations on the other hand, are very much context dependent, and dialog history contains useful information that...
2108.08405
title_snapshot
han21_interspeech
Coreference Augmentation for Multi-Domain Task-Oriented Dialogue State Tracking
[ "Ting Han", "Chongxuan Huang", "Wei Peng" ]
https://www.isca-archive.org/interspeech_2021/han21_interspeech.html
https://www.isca-archive.org/interspeech_2021/han21_interspeech.pdf
10.21437/Interspeech.2021-1463
1259-1263
@inproceedings{han21_interspeech, title = {{Coreference Augmentation for Multi-Domain Task-Oriented Dialogue State Tracking}}, author = {Ting Han and Chongxuan Huang and Wei Peng}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1259--1263}, doi = {10.21437/Interspeech.2021-1...
Dialogue State Tracking (DST), which is the process of inferring user goals by estimating belief states given the dialogue history, plays a critical role in task-oriented dialogue systems. A coreference phenomenon observed in multi-turn conversations is not addressed by existing DST models, leading to suboptimal perfor...
2106.08723
title_snapshot
arora21_interspeech
Rethinking End-to-End Evaluation of Decomposable Tasks: A Case Study on Spoken Language Understanding
[ "Siddhant Arora", "Alissa Ostapenko", "Vijay Viswanathan", "Siddharth Dalmia", "Florian Metze", "Shinji Watanabe", "Alan W. Black" ]
https://www.isca-archive.org/interspeech_2021/arora21_interspeech.html
https://www.isca-archive.org/interspeech_2021/arora21_interspeech.pdf
10.21437/Interspeech.2021-1537
1264-1268
@inproceedings{arora21_interspeech, title = {{Rethinking End-to-End Evaluation of Decomposable Tasks: A Case Study on Spoken Language Understanding}}, author = {Siddhant Arora and Alissa Ostapenko and Vijay Viswanathan and Siddharth Dalmia and Florian Metze and Shinji Watanabe and Alan W. Black}, year ...
Decomposable tasks are complex and comprise of a hierarchy of sub-tasks. Spoken intent prediction, for example, combines automatic speech recognition and natural language understanding. Existing benchmarks, however, typically hold out examples for only the surface-level sub-task. As a result, models with similar perfor...
2106.15065
title_snapshot
sun21b_interspeech
Semantic Data Augmentation for End-to-End Mandarin Speech Recognition
[ "Jianwei Sun", "Zhiyuan Tang", "Hengxin Yin", "Wei Wang", "Xi Zhao", "Shuaijiang Zhao", "Xiaoning Lei", "Wei Zou", "Xiangang Li" ]
https://www.isca-archive.org/interspeech_2021/sun21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/sun21b_interspeech.pdf
10.21437/Interspeech.2021-1162
1269-1273
@inproceedings{sun21b_interspeech, title = {{Semantic Data Augmentation for End-to-End Mandarin Speech Recognition}}, author = {Jianwei Sun and Zhiyuan Tang and Hengxin Yin and Wei Wang and Xi Zhao and Shuaijiang Zhao and Xiaoning Lei and Wei Zou and Xiangang Li}, year = {2021}, booktitle = {{Inters...
End-to-end models have gradually become the preferred option for automatic speech recognition (ASR) applications. During the training of end-to-end ASR, data augmentation is a quite effective technique for regularizing the neural networks. This paper proposes a novel data augmentation technique based on semantic transp...
2104.12521
title_snapshot
gong21c_interspeech
Layer-Wise Fast Adaptation for End-to-End Multi-Accent Speech Recognition
[ "Xun Gong", "Yizhou Lu", "Zhikai Zhou", "Yanmin Qian" ]
https://www.isca-archive.org/interspeech_2021/gong21c_interspeech.html
https://www.isca-archive.org/interspeech_2021/gong21c_interspeech.pdf
10.21437/Interspeech.2021-1075
1274-1278
@inproceedings{gong21c_interspeech, title = {{Layer-Wise Fast Adaptation for End-to-End Multi-Accent Speech Recognition}}, author = {Xun Gong and Yizhou Lu and Zhikai Zhou and Yanmin Qian}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1274--1278}, doi = {10.21437/Interspee...
Accent variability has posed a huge challenge to automatic speech recognition (ASR) modeling. Although one-hot accent vector based adaptation systems are commonly used, they require prior knowledge about the target accent and cannot handle unseen accents. Furthermore, simply concatenating accent embeddings does not mak...
2204.09883
title_snapshot
wang21m_interspeech
Low Resource German ASR with Untranscribed Data Spoken by Non-Native Children — INTERSPEECH 2021 Shared Task SPAPL System
[ "Jinhan Wang", "Yunzheng Zhu", "Ruchao Fan", "Wei Chu", "Abeer Alwan" ]
https://www.isca-archive.org/interspeech_2021/wang21m_interspeech.html
https://www.isca-archive.org/interspeech_2021/wang21m_interspeech.pdf
10.21437/Interspeech.2021-1974
1279-1283
@inproceedings{wang21m_interspeech, title = {{Low Resource German ASR with Untranscribed Data Spoken by Non-Native Children — INTERSPEECH 2021 Shared Task SPAPL System}}, author = {Jinhan Wang and Yunzheng Zhu and Ruchao Fan and Wei Chu and Abeer Alwan}, year = {2021}, booktitle = {{Interspeech 2021...
This paper describes the SPAPL system for the INTERSPEECH 2021 Challenge: Shared Task on Automatic Speech Recognition for Non-Native Children’s Speech in German. ~5 hours of transcribed data and ~60 hours of untranscribed data are provided to develop a German ASR system for children. For the training of the transcribed...
2106.09963
title_snapshot
sim21_interspeech
Robust Continuous On-Device Personalization for Automatic Speech Recognition
[ "Khe Chai Sim", "Angad Chandorkar", "Fan Gao", "Mason Chua", "Tsendsuren Munkhdalai", "Françoise Beaufays" ]
https://www.isca-archive.org/interspeech_2021/sim21_interspeech.html
https://www.isca-archive.org/interspeech_2021/sim21_interspeech.pdf
10.21437/Interspeech.2021-318
1284-1288
@inproceedings{sim21_interspeech, title = {{Robust Continuous On-Device Personalization for Automatic Speech Recognition}}, author = {Khe Chai Sim and Angad Chandorkar and Fan Gao and Mason Chua and Tsendsuren Munkhdalai and Françoise Beaufays}, year = {2021}, booktitle = {{Interspeech 2021}}, pag...
On-device personalization of an all-neural automatic speech recognition (ASR) model can be achieved efficiently by fine-tuning the last few layers of the model. This approach has been shown to be effective for adapting the model to recognize rare named entities using only a small amount of data. To reliably perform con...
null
null
kumar21b_interspeech
Speaker Normalization Using Joint Variational Autoencoder
[ "Shashi Kumar", "Shakti P. Rath", "Abhishek Pandey" ]
https://www.isca-archive.org/interspeech_2021/kumar21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/kumar21b_interspeech.pdf
10.21437/Interspeech.2021-467
1289-1293
@inproceedings{kumar21b_interspeech, title = {{Speaker Normalization Using Joint Variational Autoencoder}}, author = {Shashi Kumar and Shakti P. Rath and Abhishek Pandey}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1289--1293}, doi = {10.21437/Interspeech.2021-467}, is...
Speaker adaptation is known to provide significant improvement in speech recognition accuracy. However, in practical scenario, only a few seconds of audio is available due to which it may be infeasible to apply speaker adaptation methods such as i-vector and fMLLR robustly. Also, decoding with fMLLR transformation happ...
null
null
xu21c_interspeech
The TAL System for the INTERSPEECH2021 Shared Task on Automatic Speech Recognition for Non-Native Childrens Speech
[ "Gaopeng Xu", "Song Yang", "Lu Ma", "Chengfei Li", "Zhongqin Wu" ]
https://www.isca-archive.org/interspeech_2021/xu21c_interspeech.html
https://www.isca-archive.org/interspeech_2021/xu21c_interspeech.pdf
10.21437/Interspeech.2021-1104
1294-1298
@inproceedings{xu21c_interspeech, title = {{The TAL System for the INTERSPEECH2021 Shared Task on Automatic Speech Recognition for Non-Native Childrens Speech}}, author = {Gaopeng Xu and Song Yang and Lu Ma and Chengfei Li and Zhongqin Wu}, year = {2021}, booktitle = {{Interspeech 2021}}, pages ...
This paper describes TAL’s system for the INTERSPEECH 2021 shared task on Automatic Speech Recognition (ASR) for non-native children’s speech. In this work, we attempt to apply the self-supervised approach to non-native German children’s ASR. First, we conduct some baseline experiments to indicate that self-supervised ...
null
null
lam21b_interspeech
On-the-Fly Aligned Data Augmentation for Sequence-to-Sequence ASR
[ "Tsz Kin Lam", "Mayumi Ohta", "Shigehiko Schamoni", "Stefan Riezler" ]
https://www.isca-archive.org/interspeech_2021/lam21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/lam21b_interspeech.pdf
10.21437/Interspeech.2021-1679
1299-1303
@inproceedings{lam21b_interspeech, title = {{On-the-Fly Aligned Data Augmentation for Sequence-to-Sequence ASR}}, author = {Tsz Kin Lam and Mayumi Ohta and Shigehiko Schamoni and Stefan Riezler}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1299--1303}, doi = {10.21437/Int...
We propose an on-the-fly data augmentation method for automatic speech recognition (ASR) that uses alignment information to generate effective training samples. Our method, called Aligned Data Augmentation (ADA) for ASR, replaces transcribed tokens and the speech representations in an aligned manner to generate previou...
2104.01393
title_snapshot
gao21_interspeech
Zero-Shot Cross-Lingual Phonetic Recognition with External Language Embedding
[ "Heting Gao", "Junrui Ni", "Yang Zhang", "Kaizhi Qian", "Shiyu Chang", "Mark Hasegawa-Johnson" ]
https://www.isca-archive.org/interspeech_2021/gao21_interspeech.html
https://www.isca-archive.org/interspeech_2021/gao21_interspeech.pdf
10.21437/Interspeech.2021-1843
1304-1308
@inproceedings{gao21_interspeech, title = {{Zero-Shot Cross-Lingual Phonetic Recognition with External Language Embedding}}, author = {Heting Gao and Junrui Ni and Yang Zhang and Kaizhi Qian and Shiyu Chang and Mark Hasegawa-Johnson}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1...
Many existing languages are too sparsely resourced for monolingual deep learning networks to achieve high accuracy. Multilingual phonetic recognition systems mitigate data sparsity issues by training models on data from multiple languages and learning a speech-to-phone or speech-to-text model universal to all languages...
null
null
huang21c_interspeech
Rapid Speaker Adaptation for Conformer Transducer: Attention and Bias Are All You Need
[ "Yan Huang", "Guoli Ye", "Jinyu Li", "Yifan Gong" ]
https://www.isca-archive.org/interspeech_2021/huang21c_interspeech.html
https://www.isca-archive.org/interspeech_2021/huang21c_interspeech.pdf
10.21437/Interspeech.2021-1884
1309-1313
@inproceedings{huang21c_interspeech, title = {{Rapid Speaker Adaptation for Conformer Transducer: Attention and Bias Are All You Need}}, author = {Yan Huang and Guoli Ye and Jinyu Li and Yifan Gong}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1309--1313}, doi = {10.21437...
Conformer transducer achieves new state-of-the-art end-to-end (E2E) system performance and has become increasingly appealing for production. In this paper, we study how to effectively perform rapid speaker adaptation in a conformer transducer and how it compares with the RNN transducer. We hierarchically decompose the ...
null
null
das21b_interspeech
Best of Both Worlds: Robust Accented Speech Recognition with Adversarial Transfer Learning
[ "Nilaksh Das", "Sravan Bodapati", "Monica Sunkara", "Sundararajan Srinivasan", "Duen Horng Chau" ]
https://www.isca-archive.org/interspeech_2021/das21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/das21b_interspeech.pdf
10.21437/Interspeech.2021-1888
1314-1318
@inproceedings{das21b_interspeech, title = {{Best of Both Worlds: Robust Accented Speech Recognition with Adversarial Transfer Learning}}, author = {Nilaksh Das and Sravan Bodapati and Monica Sunkara and Sundararajan Srinivasan and Duen Horng Chau}, year = {2021}, booktitle = {{Interspeech 2021}}, ...
Training deep neural networks for automatic speech recognition (ASR) requires large amounts of transcribed speech. This becomes a bottleneck for training robust models for accented speech which typically contains high variability in pronunciation and other semantics, since obtaining large amounts of annotated accented ...
2103.05834
title_snapshot
chu21_interspeech
Extending Pronunciation Dictionary with Automatically Detected Word Mispronunciations to Improve PAII’s System for Interspeech 2021 Non-Native Child English Close Track ASR Challenge
[ "Wei Chu", "Peng Chang", "Jing Xiao" ]
https://www.isca-archive.org/interspeech_2021/chu21_interspeech.html
https://www.isca-archive.org/interspeech_2021/chu21_interspeech.pdf
10.21437/Interspeech.2021-2053
1319-1323
@inproceedings{chu21_interspeech, title = {{Extending Pronunciation Dictionary with Automatically Detected Word Mispronunciations to Improve PAII’s System for Interspeech 2021 Non-Native Child English Close Track ASR Challenge}}, author = {Wei Chu and Peng Chang and Jing Xiao}, year = {2021}, bookti...
This paper proposed to automatically detect mispronounced words over the regions that have low Goodness-of-Pronunciation scores through a constrained phone decoder, then add these word mispronunciations into the orthodox lexicon without colliding with existing pronunciations, finally use the expanded lexicon for decodi...
null
null
li21d_interspeech
CVC: Contrastive Learning for Non-Parallel Voice Conversion
[ "Tingle Li", "Yichen Liu", "Chenxu Hu", "Hang Zhao" ]
https://www.isca-archive.org/interspeech_2021/li21d_interspeech.html
https://www.isca-archive.org/interspeech_2021/li21d_interspeech.pdf
10.21437/Interspeech.2021-137
1324-1328
@inproceedings{li21d_interspeech, title = {{CVC: Contrastive Learning for Non-Parallel Voice Conversion}}, author = {Tingle Li and Yichen Liu and Chenxu Hu and Hang Zhao}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1324--1328}, doi = {10.21437/Interspeech.2021-137}, is...
Cycle consistent generative adversarial network (CycleGAN) and variational autoencoder (VAE) based models have gained popularity in non-parallel voice conversion recently. However, they often suffer from difficult training process and unsatisfactory results. In this paper, we propose a contrastive learning-based advers...
2011.00782
title_snapshot
huang21d_interspeech
A Preliminary Study of a Two-Stage Paradigm for Preserving Speaker Identity in Dysarthric Voice Conversion
[ "Wen-Chin Huang", "Kazuhiro Kobayashi", "Yu-Huai Peng", "Ching-Feng Liu", "Yu Tsao", "Hsin-Min Wang", "Tomoki Toda" ]
https://www.isca-archive.org/interspeech_2021/huang21d_interspeech.html
https://www.isca-archive.org/interspeech_2021/huang21d_interspeech.pdf
10.21437/Interspeech.2021-208
1329-1333
@inproceedings{huang21d_interspeech, title = {{A Preliminary Study of a Two-Stage Paradigm for Preserving Speaker Identity in Dysarthric Voice Conversion}}, author = {Wen-Chin Huang and Kazuhiro Kobayashi and Yu-Huai Peng and Ching-Feng Liu and Yu Tsao and Hsin-Min Wang and Tomoki Toda}, year = {2021}...
We propose a new paradigm for maintaining speaker identity in dysarthric voice conversion (DVC). The poor quality of dysarthric speech can be greatly improved by statistical VC, but as the normal speech utterances of a dysarthria patient are nearly impossible to collect, previous work failed to recover the individualit...
2106.01415
title_snapshot
eskimez21_interspeech
One-Shot Voice Conversion with Speaker-Agnostic StarGAN
[ "Sefik Emre Eskimez", "Dimitrios Dimitriadis", "Kenichi Kumatani", "Robert Gmyr" ]
https://www.isca-archive.org/interspeech_2021/eskimez21_interspeech.html
https://www.isca-archive.org/interspeech_2021/eskimez21_interspeech.pdf
10.21437/Interspeech.2021-221
1334-1338
@inproceedings{eskimez21_interspeech, title = {{One-Shot Voice Conversion with Speaker-Agnostic StarGAN}}, author = {Sefik Emre Eskimez and Dimitrios Dimitriadis and Kenichi Kumatani and Robert Gmyr}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1334--1338}, doi = {10.2143...
In this work, we propose a variant of STARGAN for many-to-many voice conversion (VC) conditioned on the d-vectors for short-duration (2–15 seconds) speech. We make several modifications to the STARGAN training and employ new network architectures. We employ a transformer encoder in the discriminator network, and we app...
null
null
koshizuka21_interspeech
Fine-Tuning Pre-Trained Voice Conversion Model for Adding New Target Speakers with Limited Data
[ "Takeshi Koshizuka", "Hidefumi Ohmura", "Kouichi Katsurada" ]
https://www.isca-archive.org/interspeech_2021/koshizuka21_interspeech.html
https://www.isca-archive.org/interspeech_2021/koshizuka21_interspeech.pdf
10.21437/Interspeech.2021-244
1339-1343
@inproceedings{koshizuka21_interspeech, title = {{Fine-Tuning Pre-Trained Voice Conversion Model for Adding New Target Speakers with Limited Data}}, author = {Takeshi Koshizuka and Hidefumi Ohmura and Kouichi Katsurada}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1339--1343}, ...
Voice conversion (VC) is a technique that converts speaker-dependent non-linguistic information into that of another speaker, while retaining the linguistic information of the input speech. A typical VC system comprises two modules: an encoder module that removes speaker individuality from the input speech and a decode...
null
null
wang21n_interspeech
VQMIVC: Vector Quantization and Mutual Information-Based Unsupervised Speech Representation Disentanglement for One-Shot Voice Conversion
[ "Disong Wang", "Liqun Deng", "Yu Ting Yeung", "Xiao Chen", "Xunying Liu", "Helen Meng" ]
https://www.isca-archive.org/interspeech_2021/wang21n_interspeech.html
https://www.isca-archive.org/interspeech_2021/wang21n_interspeech.pdf
10.21437/Interspeech.2021-283
1344-1348
@inproceedings{wang21n_interspeech, title = {{VQMIVC: Vector Quantization and Mutual Information-Based Unsupervised Speech Representation Disentanglement for One-Shot Voice Conversion}}, author = {Disong Wang and Liqun Deng and Yu Ting Yeung and Xiao Chen and Xunying Liu and Helen Meng}, year = {2021}...
One-shot voice conversion (VC), which performs conversion across arbitrary speakers with only a single target-speaker utterance for reference, can be effectively achieved by speech representation disentanglement. Existing work generally ignores the correlation between different speech representations during training, w...
2106.10132
title_snapshot
li21e_interspeech
StarGANv2-VC: A Diverse, Unsupervised, Non-Parallel Framework for Natural-Sounding Voice Conversion
[ "Yinghao Aaron Li", "Ali Zare", "Nima Mesgarani" ]
https://www.isca-archive.org/interspeech_2021/li21e_interspeech.html
https://www.isca-archive.org/interspeech_2021/li21e_interspeech.pdf
10.21437/Interspeech.2021-319
1349-1353
@inproceedings{li21e_interspeech, title = {{StarGANv2-VC: A Diverse, Unsupervised, Non-Parallel Framework for Natural-Sounding Voice Conversion}}, author = {Yinghao Aaron Li and Ali Zare and Nima Mesgarani}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1349--1353}, doi = {...
We present an unsupervised non-parallel many-to-many voice conversion (VC) method using a generative adversarial network (GAN) called StarGAN v2. Using a combination of adversarial source classifier loss and perceptual loss, our model significantly outperforms previous VC models. Although our model is trained only with...
2107.10394
title_snapshot
kumar21c_interspeech
Normalization Driven Zero-Shot Multi-Speaker Speech Synthesis
[ "Neeraj Kumar", "Srishti Goel", "Ankur Narang", "Brejesh Lall" ]
https://www.isca-archive.org/interspeech_2021/kumar21c_interspeech.html
https://www.isca-archive.org/interspeech_2021/kumar21c_interspeech.pdf
10.21437/Interspeech.2021-441
1354-1358
@inproceedings{kumar21c_interspeech, title = {{Normalization Driven Zero-Shot Multi-Speaker Speech Synthesis}}, author = {Neeraj Kumar and Srishti Goel and Ankur Narang and Brejesh Lall}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1354--1358}, doi = {10.21437/Interspeech...
In this paper, we present a novel zero-shot multi-speaker speech synthesis approach (ZSM-SS) that leverages the normalization architecture and speaker encoder with non-autoregressive multi-head attention driven encoder-decoder architecture. Given an input text and a reference speech sample of an unseen person, ZSM-SS c...
2012.07252
title_judge
sakamoto21_interspeech
StarGAN-VC+ASR: StarGAN-Based Non-Parallel Voice Conversion Regularized by Automatic Speech Recognition
[ "Shoki Sakamoto", "Akira Taniguchi", "Tadahiro Taniguchi", "Hirokazu Kameoka" ]
https://www.isca-archive.org/interspeech_2021/sakamoto21_interspeech.html
https://www.isca-archive.org/interspeech_2021/sakamoto21_interspeech.pdf
10.21437/Interspeech.2021-492
1359-1363
@inproceedings{sakamoto21_interspeech, title = {{StarGAN-VC+ASR: StarGAN-Based Non-Parallel Voice Conversion Regularized by Automatic Speech Recognition}}, author = {Shoki Sakamoto and Akira Taniguchi and Tadahiro Taniguchi and Hirokazu Kameoka}, year = {2021}, booktitle = {{Interspeech 2021}}, pa...
Preserving the linguistic content of input speech is essential during voice conversion (VC). The star generative adversarial network-based VC method (StarGAN-VC) is a recently developed method that allows non-parallel many-to-many VC. Although this method is powerful, it can fail to preserve the linguistic content of i...
2108.04395
title_snapshot
xu21d_interspeech
Two-Pathway Style Embedding for Arbitrary Voice Conversion
[ "Xuexin Xu", "Liang Shi", "Jinhui Chen", "Xunquan Chen", "Jie Lian", "Pingyuan Lin", "Zhihong Zhang", "Edwin R. Hancock" ]
https://www.isca-archive.org/interspeech_2021/xu21d_interspeech.html
https://www.isca-archive.org/interspeech_2021/xu21d_interspeech.pdf
10.21437/Interspeech.2021-506
1364-1368
@inproceedings{xu21d_interspeech, title = {{Two-Pathway Style Embedding for Arbitrary Voice Conversion}}, author = {Xuexin Xu and Liang Shi and Jinhui Chen and Xunquan Chen and Jie Lian and Pingyuan Lin and Zhihong Zhang and Edwin R. Hancock}, year = {2021}, booktitle = {{Interspeech 2021}}, pages...
Arbitrary voice conversion, also referred to as zero-shot voice conversion, has recently attracted increased attention in the literature. Although disentangling the linguistic and style representations for acoustic features is an effective way to achieve zero-shot voice conversion, the problem of how to convert to a na...
null
null
liu21c_interspeech
Non-Parallel Any-to-Many Voice Conversion by Replacing Speaker Statistics
[ "Yufei Liu", "Chengzhu Yu", "Wang Shuai", "Zhenchuan Yang", "Yang Chao", "Weibin Zhang" ]
https://www.isca-archive.org/interspeech_2021/liu21c_interspeech.html
https://www.isca-archive.org/interspeech_2021/liu21c_interspeech.pdf
10.21437/Interspeech.2021-557
1369-1373
@inproceedings{liu21c_interspeech, title = {{Non-Parallel Any-to-Many Voice Conversion by Replacing Speaker Statistics}}, author = {Yufei Liu and Chengzhu Yu and Wang Shuai and Zhenchuan Yang and Yang Chao and Weibin Zhang}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1369--1373}...
This paper proposes a non-parallel any-to-many voice conversion (VC) approach with a novel statistics replacement layer. Non-parallel VC is usually achieved by firstly disentangling linguistic and speaker representations, and then concatenating the linguistic content with the learned target speaker’s embedding at the c...
null
null
zhou21c_interspeech
Cross-Lingual Voice Conversion with a Cycle Consistency Loss on Linguistic Representation
[ "Yi Zhou", "Xiaohai Tian", "Zhizheng Wu", "Haizhou Li" ]
https://www.isca-archive.org/interspeech_2021/zhou21c_interspeech.html
https://www.isca-archive.org/interspeech_2021/zhou21c_interspeech.pdf
10.21437/Interspeech.2021-687
1374-1378
@inproceedings{zhou21c_interspeech, title = {{Cross-Lingual Voice Conversion with a Cycle Consistency Loss on Linguistic Representation}}, author = {Yi Zhou and Xiaohai Tian and Zhizheng Wu and Haizhou Li}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1374--1378}, doi = {1...
Cross-Lingual Voice Conversion (XVC) aims to modify a source speaker identity towards a target while preserving the source linguistic content. This paper introduces a cycle consistency loss on linguistic representation to ensure the speech content unchanged after conversion. The proposed XVC model consists of two loss ...
null
null
du21_interspeech
Improving Robustness of One-Shot Voice Conversion with Deep Discriminative Speaker Encoder
[ "Hongqiang Du", "Lei Xie" ]
https://www.isca-archive.org/interspeech_2021/du21_interspeech.html
https://www.isca-archive.org/interspeech_2021/du21_interspeech.pdf
10.21437/Interspeech.2021-2132
1379-1383
@inproceedings{du21_interspeech, title = {{Improving Robustness of One-Shot Voice Conversion with Deep Discriminative Speaker Encoder}}, author = {Hongqiang Du and Lei Xie}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1379--1383}, doi = {10.21437/Interspeech.2021-2132}, ...
One-shot voice conversion has received significant attention since only one utterance from source speaker and target speaker respectively is required. Moreover, source speaker and target speaker do not need to be seen during training. However, available one-shot voice conversion approaches are not stable for unseen spe...
2106.10406
title_snapshot
white21_interspeech
Optimizing an Automatic Creaky Voice Detection Method for Australian English Speaking Females
[ "Hannah White", "Joshua Penney", "Andy Gibson", "Anita Szakay", "Felicity Cox" ]
https://www.isca-archive.org/interspeech_2021/white21_interspeech.html
https://www.isca-archive.org/interspeech_2021/white21_interspeech.pdf
10.21437/Interspeech.2021-711
1384-1388
@inproceedings{white21_interspeech, title = {{Optimizing an Automatic Creaky Voice Detection Method for Australian English Speaking Females}}, author = {Hannah White and Joshua Penney and Andy Gibson and Anita Szakay and Felicity Cox}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {...
Creaky voice is a nonmodal phonation type that has various linguistic and sociolinguistic functions. Manually annotating creaky voice for phonetic analysis is time-consuming and labor-intensive. In recent years, automatic tools for detecting creaky voice have been proposed, which present the possibility for easier, fas...
null
null
penney21_interspeech
A Comparison of Acoustic Correlates of Voice Quality Across Different Recording Devices: A Cautionary Tale
[ "Joshua Penney", "Andy Gibson", "Felicity Cox", "Michael Proctor", "Anita Szakay" ]
https://www.isca-archive.org/interspeech_2021/penney21_interspeech.html
https://www.isca-archive.org/interspeech_2021/penney21_interspeech.pdf
10.21437/Interspeech.2021-729
1389-1393
@inproceedings{penney21_interspeech, title = {{A Comparison of Acoustic Correlates of Voice Quality Across Different Recording Devices: A Cautionary Tale}}, author = {Joshua Penney and Andy Gibson and Felicity Cox and Michael Proctor and Anita Szakay}, year = {2021}, booktitle = {{Interspeech 2021}}...
There has been a recent increase in speech research utilizing data recorded with participants’ personal devices, particularly in light of the COVID-19 pandemic and restrictions on face-to-face interactions. This raises important questions about whether these recordings are comparable to those made in traditional lab-ba...
null
null
sfakianaki21_interspeech
Investigating Voice Function Characteristics of Greek Speakers with Hearing Loss Using Automatic Glottal Source Feature Extraction
[ "Anna Sfakianaki", "George P. Kafentzis" ]
https://www.isca-archive.org/interspeech_2021/sfakianaki21_interspeech.html
https://www.isca-archive.org/interspeech_2021/sfakianaki21_interspeech.pdf
10.21437/Interspeech.2021-870
1394-1398
@inproceedings{sfakianaki21_interspeech, title = {{Investigating Voice Function Characteristics of Greek Speakers with Hearing Loss Using Automatic Glottal Source Feature Extraction}}, author = {Anna Sfakianaki and George P. Kafentzis}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = ...
The current study investigates voice quality characteristics of Greek adults with normal hearing and hearing loss, automatically obtained from glottal inverse filtering analysis using the Aalto Aparat toolkit. Aalto Aparat has been employed in glottal flow analysis of disordered speech, but to the best of the authors’ ...
null
null
huckvale21_interspeech
Automated Detection of Voice Disorder in the Saarbrücken Voice Database: Effects of Pathology Subset and Audio Materials
[ "Mark Huckvale", "Catinca Buciuleac" ]
https://www.isca-archive.org/interspeech_2021/huckvale21_interspeech.html
https://www.isca-archive.org/interspeech_2021/huckvale21_interspeech.pdf
10.21437/Interspeech.2021-1507
1399-1403
@inproceedings{huckvale21_interspeech, title = {{Automated Detection of Voice Disorder in the Saarbrücken Voice Database: Effects of Pathology Subset and Audio Materials}}, author = {Mark Huckvale and Catinca Buciuleac}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1399--1403}, ...
The Saarbrücken Voice Database contains speech and simultaneous electroglottography recordings of 1002 speakers exhibiting a wide range of voice disorders, together with recordings of 851 controls. Previous studies have used this database to build systems for automated detection of voice disorders and for differential ...
null
null
lulich21_interspeech
Accelerometer-Based Measurements of Voice Quality in Children During Semi-Occluded Vocal Tract Exercise with a Narrow Straw in Air
[ "Steven M. Lulich", "Rita R. Patel" ]
https://www.isca-archive.org/interspeech_2021/lulich21_interspeech.html
https://www.isca-archive.org/interspeech_2021/lulich21_interspeech.pdf
10.21437/Interspeech.2021-1918
1404-1408
@inproceedings{lulich21_interspeech, title = {{Accelerometer-Based Measurements of Voice Quality in Children During Semi-Occluded Vocal Tract Exercise with a Narrow Straw in Air}}, author = {Steven M. Lulich and Rita R. Patel}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1404--14...
Non-invasive measures of voice quality, such as H1-H2, rely on oral flow signals, inverse filtered speech signals, or corrections for the effects of formants. Voice quality measures play especially important roles in the assessment of voice disorders and the evaluation of treatment efficacy. One type of treatment that ...
null
null
perez21_interspeech
Articulatory Coordination for Speech Motor Tracking in Huntington Disease
[ "Matthew Perez", "Amrit Romana", "Angela Roberts", "Noelle Carlozzi", "Jennifer Ann Miner", "Praveen Dayalu", "Emily Mower Provost" ]
https://www.isca-archive.org/interspeech_2021/perez21_interspeech.html
https://www.isca-archive.org/interspeech_2021/perez21_interspeech.pdf
10.21437/Interspeech.2021-688
1409-1413
@inproceedings{perez21_interspeech, title = {{Articulatory Coordination for Speech Motor Tracking in Huntington Disease}}, author = {Matthew Perez and Amrit Romana and Angela Roberts and Noelle Carlozzi and Jennifer Ann Miner and Praveen Dayalu and Emily Mower Provost}, year = {2021}, booktitle = {{...
Huntington Disease (HD) is a progressive disorder which often manifests in motor impairment. Motor severity (captured via motor score) is a key component in assessing overall HD severity. However, motor score evaluation involves in-clinic visits with a trained medical professional, which are expensive and not always ac...
2109.13815
title_snapshot
ferrer21_interspeech
Modeling Dysphonia Severity as a Function of Roughness and Breathiness Ratings in the GRBAS Scale
[ "Carlos A. Ferrer", "Efren Aragón", "María E. Hdez-Díaz", "Marc S. de Bodt", "Roman Cmejla", "Marina Englert", "Mara Behlau", "Elmar Nöth" ]
https://www.isca-archive.org/interspeech_2021/ferrer21_interspeech.html
https://www.isca-archive.org/interspeech_2021/ferrer21_interspeech.pdf
10.21437/Interspeech.2021-1540
1414-1418
@inproceedings{ferrer21_interspeech, title = {{Modeling Dysphonia Severity as a Function of Roughness and Breathiness Ratings in the GRBAS Scale}}, author = {Carlos A. Ferrer and Efren Aragón and María E. Hdez-Díaz and Marc S. de Bodt and Roman Cmejla and Marina Englert and Mara Behlau and Elmar Nöth}, yea...
Dysphonia comprises many perceptually deviating aspects of voice, and its overall severity perception is made by the listener according to methods of aggregating the single dimensions which are personally conceived and not well studied. Roughness and breathiness are constituent dimensions in most devised rating scales ...
null
null
karpov21_interspeech
Golos: Russian Dataset for Speech Research
[ "Nikolay Karpov", "Alexander Denisenko", "Fedor Minkin" ]
https://www.isca-archive.org/interspeech_2021/karpov21_interspeech.html
https://www.isca-archive.org/interspeech_2021/karpov21_interspeech.pdf
10.21437/Interspeech.2021-462
1419-1423
@inproceedings{karpov21_interspeech, title = {{Golos: Russian Dataset for Speech Research}}, author = {Nikolay Karpov and Alexander Denisenko and Fedor Minkin}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1419--1423}, doi = {10.21437/Interspeech.2021-462}, issn = {...
This paper introduces a novel Russian speech dataset called Golos, a large corpus suitable for speech research. The dataset mainly consists of recorded audio files manually annotated on the crowd-sourcing platform. The total duration of the audio is about 1240 hours. We have made the corpus freely available to download...
2106.10161
title_snapshot
sadhu21b_interspeech
Radically Old Way of Computing Spectra: Applications in End-to-End ASR
[ "Samik Sadhu", "Hynek Hermansky" ]
https://www.isca-archive.org/interspeech_2021/sadhu21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/sadhu21b_interspeech.pdf
10.21437/Interspeech.2021-643
1424-1428
@inproceedings{sadhu21b_interspeech, title = {{Radically Old Way of Computing Spectra: Applications in End-to-End ASR}}, author = {Samik Sadhu and Hynek Hermansky}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1424--1428}, doi = {10.21437/Interspeech.2021-643}, issn ...
We propose a technique to compute spectrograms using Frequency Domain Linear Prediction (FDLP) that uses all-pole models to fit the squared Hilbert envelope of speech in different frequency sub-bands. The spectrogram of a complete speech utterance is computed by overlap-add of contiguous all-pole model responses. A lon...
2103.14129
title_snapshot
alghezi21_interspeech
Self-Supervised End-to-End ASR for Low Resource L2 Swedish
[ "Ragheb Al-Ghezi", "Yaroslav Getman", "Aku Rouhe", "Raili Hildén", "Mikko Kurimo" ]
https://www.isca-archive.org/interspeech_2021/alghezi21_interspeech.html
https://www.isca-archive.org/interspeech_2021/alghezi21_interspeech.pdf
10.21437/Interspeech.2021-1710
1429-1433
@inproceedings{alghezi21_interspeech, title = {{Self-Supervised End-to-End ASR for Low Resource L2 Swedish}}, author = {Ragheb Al-Ghezi and Yaroslav Getman and Aku Rouhe and Raili Hildén and Mikko Kurimo}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1429--1433}, doi = {10...
Unlike traditional (hybrid) Automatic Speech Recognition (ASR), end-to-end ASR systems simplify the training procedure by directly mapping acoustic features to sequences of graphemes or characters, thereby eliminating the need for specialized acoustic, language, or pronunciation models. However, one drawback of end-to-...
null
null
oneill21_interspeech
SPGISpeech: 5,000 Hours of Transcribed Financial Audio for Fully Formatted End-to-End Speech Recognition
[ "Patrick K. O’Neill", "Vitaly Lavrukhin", "Somshubra Majumdar", "Vahid Noroozi", "Yuekai Zhang", "Oleksii Kuchaiev", "Jagadeesh Balam", "Yuliya Dovzhenko", "Keenan Freyberg", "Michael D. Shulman", "Boris Ginsburg", "Shinji Watanabe", "Georg Kucsko" ]
https://www.isca-archive.org/interspeech_2021/oneill21_interspeech.html
https://www.isca-archive.org/interspeech_2021/oneill21_interspeech.pdf
10.21437/Interspeech.2021-1860
1434-1438
@inproceedings{oneill21_interspeech, title = {{SPGISpeech: 5,000 Hours of Transcribed Financial Audio for Fully Formatted End-to-End Speech Recognition}}, author = {Patrick K. O’Neill and Vitaly Lavrukhin and Somshubra Majumdar and Vahid Noroozi and Yuekai Zhang and Oleksii Kuchaiev and Jagadeesh Balam and Y...
In the English speech-to-text (STT) machine learning task, acoustic models are conventionally trained on uncased Latin characters, and any necessary orthography (such as capitalization, punctuation, and denormalization of non-standard words) is imputed by separate post-processing models. This adds complexity and limits...
2104.02014
title_snapshot
evain21_interspeech
: A Reproducible Framework for Assessing Self-Supervised Representation Learning from Speech
[ "Solène Evain", "Ha Nguyen", "Hang Le", "Marcely Zanon Boito", "Salima Mdhaffar", "Sina Alisamir", "Ziyi Tong", "Natalia Tomashenko", "Marco Dinarelli", "Titouan Parcollet", "Alexandre Allauzen", "Yannick Estève", "Benjamin Lecouteux", "François Portet", "Solange Rossato", "Fabien Ring...
https://www.isca-archive.org/interspeech_2021/evain21_interspeech.html
https://www.isca-archive.org/interspeech_2021/evain21_interspeech.pdf
10.21437/Interspeech.2021-556
1439-1443
@inproceedings{evain21_interspeech, title = {{ LeBenchmark: A Reproducible Framework for Assessing Self-Supervised Representation Learning from Speech}}, author = {Solène Evain and Ha Nguyen and Hang Le and Marcely Zanon Boito and Salima Mdhaffar and Sina Alisamir and Ziyi Tong and Natalia Tomashenko and Mar...
Self-Supervised Learning (SSL) using huge unlabeled data has been successfully explored for image and natural language processing. Recent works also investigated SSL from speech. They were notably successful to improve performance on downstream tasks such as automatic speech recognition (ASR). While these works suggest...
2104.11462
title_judge
sturm21_interspeech
Prosodic Accommodation in Face-to-Face and Telephone Dialogues
[ "Pavel Šturm", "Radek Skarnitzl", "Tomáš Nechanský" ]
https://www.isca-archive.org/interspeech_2021/sturm21_interspeech.html
https://www.isca-archive.org/interspeech_2021/sturm21_interspeech.pdf
10.21437/Interspeech.2021-130
1444-1448
@inproceedings{sturm21_interspeech, title = {{Prosodic Accommodation in Face-to-Face and Telephone Dialogues}}, author = {Pavel Šturm and Radek Skarnitzl and Tomáš Nechanský}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1444--1448}, doi = {10.21437/Interspeech.2021-130}, ...
The study of phonetic accommodation in various communicative situations is still relatively limited. This paper examines accommodation in spontaneous conversations of eight pairs of Czech young male speakers in two communicative conditions: unconstrained face-to-face conversation and goal-oriented interaction via mobil...
null
null
riverincoutlee21_interspeech
Dialect Features in Heterogeneous and Homogeneous Gheg Speaking Communities
[ "Josiane Riverin-Coutlée", "Conceição Cunha", "Enkeleida Kapia", "Jonathan Harrington" ]
https://www.isca-archive.org/interspeech_2021/riverincoutlee21_interspeech.html
https://www.isca-archive.org/interspeech_2021/riverincoutlee21_interspeech.pdf
10.21437/Interspeech.2021-1090
1449-1453
@inproceedings{riverincoutlee21_interspeech, title = {{Dialect Features in Heterogeneous and Homogeneous Gheg Speaking Communities}}, author = {Josiane Riverin-Coutlée and Conceição Cunha and Enkeleida Kapia and Jonathan Harrington}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {14...
This apparent and real time study analyses how dialect features in the speech of children and adults are differently affected depending on whether they live in homogeneous or heterogeneous speech communities. The general hypotheses are that speakers in such high contact settings as heterogeneous urban centers are more ...
null
null
zellers21_interspeech
An Exploration of the Acoustic Space of Rhotics and Laterals in Ruruuli
[ "Margaret Zellers", "Alena Witzlack-Makarevich", "Lilja Saeboe", "Saudah Namyalo" ]
https://www.isca-archive.org/interspeech_2021/zellers21_interspeech.html
https://www.isca-archive.org/interspeech_2021/zellers21_interspeech.pdf
10.21437/Interspeech.2021-1328
1454-1458
@inproceedings{zellers21_interspeech, title = {{An Exploration of the Acoustic Space of Rhotics and Laterals in Ruruuli}}, author = {Margaret Zellers and Alena Witzlack-Makarevich and Lilja Saeboe and Saudah Namyalo}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1454--1458}, doi...
Liquid consonants — rhotics and laterals — have been shown to demonstrate unique distributional patterns cross-linguistically. It is also claimed that rhotics are more difficult to distinguish from one another phonetically than laterals, and that rhotics are less flexible than laterals when it comes to participation in...
null
null
bodur21_interspeech
Domain-Initial Strengthening in Turkish: Acoustic Cues to Prosodic Hierarchy in Stop Consonants
[ "Kubra Bodur", "Sweeney Branje", "Morgane Peirolo", "Ingrid Tiscareno", "James S. German" ]
https://www.isca-archive.org/interspeech_2021/bodur21_interspeech.html
https://www.isca-archive.org/interspeech_2021/bodur21_interspeech.pdf
10.21437/Interspeech.2021-2230
1459-1463
@inproceedings{bodur21_interspeech, title = {{Domain-Initial Strengthening in Turkish: Acoustic Cues to Prosodic Hierarchy in Stop Consonants}}, author = {Kubra Bodur and Sweeney Branje and Morgane Peirolo and Ingrid Tiscareno and James S. German}, year = {2021}, booktitle = {{Interspeech 2021}}, ...
Studies have shown that cross-linguistically, consonants at the left edge of higher-level prosodic boundaries tend to be more forcefully articulated than those at lower-level boundaries, a phenomenon known as domain-initial strengthening . This study tests whether similar effects occur in Turkish, using the Autosegment...
null
null
zmolikova21_interspeech
Auxiliary Loss Function for Target Speech Extraction and Recognition with Weak Supervision Based on Speaker Characteristics
[ "Katerina Zmolikova", "Marc Delcroix", "Desh Raj", "Shinji Watanabe", "Jan Černocký" ]
https://www.isca-archive.org/interspeech_2021/zmolikova21_interspeech.html
https://www.isca-archive.org/interspeech_2021/zmolikova21_interspeech.pdf
10.21437/Interspeech.2021-986
1464-1468
@inproceedings{zmolikova21_interspeech, title = {{Auxiliary Loss Function for Target Speech Extraction and Recognition with Weak Supervision Based on Speaker Characteristics}}, author = {Katerina Zmolikova and Marc Delcroix and Desh Raj and Shinji Watanabe and Jan Černocký}, year = {2021}, booktitle...
Automatic speech recognition systems deteriorate in presence of overlapped speech. A popular approach to alleviate this is target speech extraction. The extraction system is usually trained with a loss function measuring the discrepancy between the estimated and the reference target speech. This often leads to distorti...
null
null
borsdorf21_interspeech
Universal Speaker Extraction in the Presence and Absence of Target Speakers for Speech of One and Two Talkers
[ "Marvin Borsdorf", "Chenglin Xu", "Haizhou Li", "Tanja Schultz" ]
https://www.isca-archive.org/interspeech_2021/borsdorf21_interspeech.html
https://www.isca-archive.org/interspeech_2021/borsdorf21_interspeech.pdf
10.21437/Interspeech.2021-1939
1469-1473
@inproceedings{borsdorf21_interspeech, title = {{Universal Speaker Extraction in the Presence and Absence of Target Speakers for Speech of One and Two Talkers}}, author = {Marvin Borsdorf and Chenglin Xu and Haizhou Li and Tanja Schultz}, year = {2021}, booktitle = {{Interspeech 2021}}, pages ...
Speaker extraction has been studied mostly for the scenarios where a target speaker is present in a two or more talkers mixture. Such scenarios do not adequately reflect everyday conversations. For example, a target speaker can be the only active talker, be quiet for a while, or leave the conversation, that means the t...
null
null
mateju21_interspeech
Using X-Vectors for Speech Activity Detection in Broadcast Streams
[ "Lukas Mateju", "Frantisek Kynych", "Petr Cerva", "Jindrich Zdansky", "Jiri Malek" ]
https://www.isca-archive.org/interspeech_2021/mateju21_interspeech.html
https://www.isca-archive.org/interspeech_2021/mateju21_interspeech.pdf
10.21437/Interspeech.2021-192
1474-1478
@inproceedings{mateju21_interspeech, title = {{Using X-Vectors for Speech Activity Detection in Broadcast Streams}}, author = {Lukas Mateju and Frantisek Kynych and Petr Cerva and Jindrich Zdansky and Jiri Malek}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1474--1478}, doi ...
A new approach to speech activity detection (SAD) is presented in this work. It allows us to reduce the complexity and computation demands, namely in services that process streaming speech, where a SAD module usually forms the first block of the data pipeline (e.g., in a platform for 24/7 broadcast transcription). Our ...
null
null
salvati21_interspeech
Time Delay Estimation for Speaker Localization Using CNN-Based Parametrized GCC-PHAT Features
[ "Daniele Salvati", "Carlo Drioli", "Gian Luca Foresti" ]
https://www.isca-archive.org/interspeech_2021/salvati21_interspeech.html
https://www.isca-archive.org/interspeech_2021/salvati21_interspeech.pdf
10.21437/Interspeech.2021-988
1479-1483
@inproceedings{salvati21_interspeech, title = {{Time Delay Estimation for Speaker Localization Using CNN-Based Parametrized GCC-PHAT Features}}, author = {Daniele Salvati and Carlo Drioli and Gian Luca Foresti}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1479--1483}, doi ...
We propose a time delay estimation (TDE) method for speaker localization based on parametrized generalized cross-correlation phase transform (PGCC-PHAT) functions and convolutional neural networks (CNNs). The PGCC-PHAT is used to build a feature matrix, which gives TDE information of two microphone signals with differe...
null
null
yousefi21_interspeech
Real-Time Speaker Counting in a Cocktail Party Scenario Using Attention-Guided Convolutional Neural Network
[ "Midia Yousefi", "John H.L. Hansen" ]
https://www.isca-archive.org/interspeech_2021/yousefi21_interspeech.html
https://www.isca-archive.org/interspeech_2021/yousefi21_interspeech.pdf
10.21437/Interspeech.2021-331
1484-1488
@inproceedings{yousefi21_interspeech, title = {{Real-Time Speaker Counting in a Cocktail Party Scenario Using Attention-Guided Convolutional Neural Network}}, author = {Midia Yousefi and John H.L. Hansen}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1484--1488}, doi = {10...
Most current speech technology systems are designed to operate well even in the presence of multiple active speakers. However, most solutions assume that the number of co-current speakers is known. Unfortunately, this information might not always be available in real-world applications. In this study, we propose a real...
2111.00316
title_snapshot
liu21d_interspeech
End-to-End Language Diarization for Bilingual Code-Switching Speech
[ "Hexin Liu", "Leibny Paola García Perera", "Xinyi Zhang", "Justin Dauwels", "Andy W.H. Khong", "Sanjeev Khudanpur", "Suzy J. Styles" ]
https://www.isca-archive.org/interspeech_2021/liu21d_interspeech.html
https://www.isca-archive.org/interspeech_2021/liu21d_interspeech.pdf
10.21437/Interspeech.2021-82
1489-1493
@inproceedings{liu21d_interspeech, title = {{End-to-End Language Diarization for Bilingual Code-Switching Speech}}, author = {Hexin Liu and Leibny Paola García Perera and Xinyi Zhang and Justin Dauwels and Andy W.H. Khong and Sanjeev Khudanpur and Suzy J. Styles}, year = {2021}, booktitle = {{Inters...
We propose two end-to-end neural configurations for language diarization on bilingual code-switching speech. The first, a BLSTM-E2E architecture, includes a set of stacked bidirectional LSTMs to compute embeddings and incorporates the deep clustering loss to enforce grouping of languages belonging to the same class. Th...
null
null
duroselle21_interspeech
Modeling and Training Strategies for Language Recognition Systems
[ "Raphaël Duroselle", "Md. Sahidullah", "Denis Jouvet", "Irina Illina" ]
https://www.isca-archive.org/interspeech_2021/duroselle21_interspeech.html
https://www.isca-archive.org/interspeech_2021/duroselle21_interspeech.pdf
10.21437/Interspeech.2021-277
1494-1498
@inproceedings{duroselle21_interspeech, title = {{Modeling and Training Strategies for Language Recognition Systems}}, author = {Raphaël Duroselle and Md. Sahidullah and Denis Jouvet and Irina Illina}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1494--1498}, doi = {10.214...
Automatic speech recognition is complementary to language recognition. The language recognition systems exploit this complementarity by using frame-level bottleneck features extracted from neural networks trained with a phone recognition task. Recent methods apply frame-level bottleneck features extracted from an end-t...
null
null
wang21o_interspeech
A Weight Moving Average Based Alternate Decoupled Learning Algorithm for Long-Tailed Language Identification
[ "Hui Wang", "Lin Liu", "Yan Song", "Lei Fang", "Ian McLoughlin", "Li-Rong Dai" ]
https://www.isca-archive.org/interspeech_2021/wang21o_interspeech.html
https://www.isca-archive.org/interspeech_2021/wang21o_interspeech.pdf
10.21437/Interspeech.2021-776
1499-1503
@inproceedings{wang21o_interspeech, title = {{A Weight Moving Average Based Alternate Decoupled Learning Algorithm for Long-Tailed Language Identification}}, author = {Hui Wang and Lin Liu and Yan Song and Lei Fang and Ian McLoughlin and Li-Rong Dai}, year = {2021}, booktitle = {{Interspeech 2021}},...
Language identification (LID) research has made tremendous progress in recent years, especially with the introduction of deep learning techniques. However, for real-world applications where the distribution of different language data is highly imbalanced, the performance of existing LID systems is still far from satisf...
null
null
deng21b_interspeech
Improving Accent Identification and Accented Speech Recognition Under a Framework of Self-Supervised Learning
[ "Keqi Deng", "Songjun Cao", "Long Ma" ]
https://www.isca-archive.org/interspeech_2021/deng21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/deng21b_interspeech.pdf
10.21437/Interspeech.2021-1186
1504-1508
@inproceedings{deng21b_interspeech, title = {{Improving Accent Identification and Accented Speech Recognition Under a Framework of Self-Supervised Learning}}, author = {Keqi Deng and Songjun Cao and Long Ma}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1504--1508}, doi = ...
Recently, self-supervised pre-training has gained success in automatic speech recognition (ASR). However, considering the difference between speech accents in real scenarios, how to identify accents and use accent features to improve ASR is still challenging. In this paper, we employ the self-supervised pre-training me...
2109.07349
title_snapshot
fan21_interspeech
Exploring wav2vec 2.0 on Speaker Verification and Language Identification
[ "Zhiyun Fan", "Meng Li", "Shiyu Zhou", "Bo Xu" ]
https://www.isca-archive.org/interspeech_2021/fan21_interspeech.html
https://www.isca-archive.org/interspeech_2021/fan21_interspeech.pdf
10.21437/Interspeech.2021-1280
1509-1513
@inproceedings{fan21_interspeech, title = {{Exploring wav2vec 2.0 on Speaker Verification and Language Identification}}, author = {Zhiyun Fan and Meng Li and Shiyu Zhou and Bo Xu}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1509--1513}, doi = {10.21437/Interspeech.2021-1...
wav2vec 2.0 is a recently proposed self-supervised framework for speech representation learning. It follows a two-stage training process of pre-training and fine-tuning, and performs well in speech recognition tasks especially ultra-low resource cases. In this work, we attempt to extend the self-supervised framework to...
2012.06185
title_snapshot