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
ramesh21_interspeech
Self-Supervised Phonotactic Representations for Language Identification
[ "G. Ramesh", "C. Shiva Kumar", "K. Sri Rama Murty" ]
https://www.isca-archive.org/interspeech_2021/ramesh21_interspeech.html
https://www.isca-archive.org/interspeech_2021/ramesh21_interspeech.pdf
10.21437/Interspeech.2021-1310
1514-1518
@inproceedings{ramesh21_interspeech, title = {{Self-Supervised Phonotactic Representations for Language Identification}}, author = {G. Ramesh and C. Shiva Kumar and K. Sri Rama Murty}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1514--1518}, doi = {10.21437/Interspeech.20...
Phonotactic constraints characterize the sequence of permissible phoneme structures in a language and hence form an important cue for language identification (LID) task. As phonotactic constraints span across multiple phonemes, the short-term spectral analysis (20–30 ms) alone is not sufficient to capture them. The spe...
null
null
zhang21j_interspeech
E2E-Based Multi-Task Learning Approach to Joint Speech and Accent Recognition
[ "Jicheng Zhang", "Yizhou Peng", "Van Tung Pham", "Haihua Xu", "Hao Huang", "Eng Siong Chng" ]
https://www.isca-archive.org/interspeech_2021/zhang21j_interspeech.html
https://www.isca-archive.org/interspeech_2021/zhang21j_interspeech.pdf
10.21437/Interspeech.2021-1495
1519-1523
@inproceedings{zhang21j_interspeech, title = {{E2E-Based Multi-Task Learning Approach to Joint Speech and Accent Recognition}}, author = {Jicheng Zhang and Yizhou Peng and Van Tung Pham and Haihua Xu and Hao Huang and Eng Siong Chng}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1...
In this paper, we propose a single multi-task learning framework to perform End-to-End (E2E) speech recognition (ASR) and accent recognition (AR) simultaneously. The proposed framework is not only more compact but can also yield comparable or even better results than standalone systems. Specifically, we found that the ...
2106.08211
title_snapshot
tzudir21_interspeech
Excitation Source Feature Based Dialect Identification in Ao — A Low Resource Language
[ "Moakala Tzudir", "Shikha Baghel", "Priyankoo Sarmah", "S.R. Mahadeva Prasanna" ]
https://www.isca-archive.org/interspeech_2021/tzudir21_interspeech.html
https://www.isca-archive.org/interspeech_2021/tzudir21_interspeech.pdf
10.21437/Interspeech.2021-1672
1524-1528
@inproceedings{tzudir21_interspeech, title = {{Excitation Source Feature Based Dialect Identification in Ao — A Low Resource Language}}, author = {Moakala Tzudir and Shikha Baghel and Priyankoo Sarmah and S.R. Mahadeva Prasanna}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1524--...
Ao is an under-resourced Tibeto-Burman tonal language spoken in Nagaland, India. There are three distinct dialects of the language, namely, Chungli, Mongsen and Changki. The objective of dialect identification is to identify one dialect from the other within the same language family. The goal of this study is to ascert...
null
null
khare21_interspeech
Low Resource ASR: The Surprising Effectiveness of High Resource Transliteration
[ "Shreya Khare", "Ashish Mittal", "Anuj Diwan", "Sunita Sarawagi", "Preethi Jyothi", "Samarth Bharadwaj" ]
https://www.isca-archive.org/interspeech_2021/khare21_interspeech.html
https://www.isca-archive.org/interspeech_2021/khare21_interspeech.pdf
10.21437/Interspeech.2021-2062
1529-1533
@inproceedings{khare21_interspeech, title = {{Low Resource ASR: The Surprising Effectiveness of High Resource Transliteration}}, author = {Shreya Khare and Ashish Mittal and Anuj Diwan and Sunita Sarawagi and Preethi Jyothi and Samarth Bharadwaj}, year = {2021}, booktitle = {{Interspeech 2021}}, p...
Cross-lingual transfer of knowledge from high-resource languages to low-resource languages is an important research problem in automatic speech recognition (ASR). We propose a new strategy of transfer learning by pretraining using large amounts of speech in the high-resource language but with its text transliterated to...
null
null
feng21_interspeech
Unsupervised Acoustic Unit Discovery by Leveraging a Language-Independent Subword Discriminative Feature Representation
[ "Siyuan Feng", "Piotr Żelasko", "Laureano Moro-Velázquez", "Odette Scharenborg" ]
https://www.isca-archive.org/interspeech_2021/feng21_interspeech.html
https://www.isca-archive.org/interspeech_2021/feng21_interspeech.pdf
10.21437/Interspeech.2021-1664
1534-1538
@inproceedings{feng21_interspeech, title = {{Unsupervised Acoustic Unit Discovery by Leveraging a Language-Independent Subword Discriminative Feature Representation}}, author = {Siyuan Feng and Piotr Żelasko and Laureano Moro-Velázquez and Odette Scharenborg}, year = {2021}, booktitle = {{Interspeec...
This paper tackles automatically discovering phone-like acoustic units (AUD) from unlabeled speech data. Past studies usually proposed single-step approaches. We propose a two-stage approach: the first stage learns a subword-discriminative feature representation, and the second stage applies clustering to the learned r...
2104.00994
title_snapshot
kamper21_interspeech
Towards Unsupervised Phone and Word Segmentation Using Self-Supervised Vector-Quantized Neural Networks
[ "Herman Kamper", "Benjamin van Niekerk" ]
https://www.isca-archive.org/interspeech_2021/kamper21_interspeech.html
https://www.isca-archive.org/interspeech_2021/kamper21_interspeech.pdf
10.21437/Interspeech.2021-50
1539-1543
@inproceedings{kamper21_interspeech, title = {{Towards Unsupervised Phone and Word Segmentation Using Self-Supervised Vector-Quantized Neural Networks}}, author = {Herman Kamper and Benjamin van Niekerk}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1539--1543}, doi = {10....
We investigate segmenting and clustering speech into low-bitrate phone-like sequences without supervision. We specifically constrain pretrained self-supervised vector-quantized (VQ) neural networks so that blocks of contiguous feature vectors are assigned to the same code, thereby giving a variable-rate segmentation of...
2012.07551
title_snapshot
jiang21_interspeech
Speech SimCLR: Combining Contrastive and Reconstruction Objective for Self-Supervised Speech Representation Learning
[ "Dongwei Jiang", "Wubo Li", "Miao Cao", "Wei Zou", "Xiangang Li" ]
https://www.isca-archive.org/interspeech_2021/jiang21_interspeech.html
https://www.isca-archive.org/interspeech_2021/jiang21_interspeech.pdf
10.21437/Interspeech.2021-391
1544-1548
@inproceedings{jiang21_interspeech, title = {{Speech SimCLR: Combining Contrastive and Reconstruction Objective for Self-Supervised Speech Representation Learning}}, author = {Dongwei Jiang and Wubo Li and Miao Cao and Wei Zou and Xiangang Li}, year = {2021}, booktitle = {{Interspeech 2021}}, page...
Self-supervised visual pretraining has shown significant progress recently. Among those methods, SimCLR greatly advanced the state of the art in self-supervised and semi-supervised learning on ImageNet. The input feature representations for speech and visual tasks are both continuous, so it is natural to consider apply...
2010.13991
title_snapshot
jacobs21_interspeech
Multilingual Transfer of Acoustic Word Embeddings Improves When Training on Languages Related to the Target Zero-Resource Language
[ "Christiaan Jacobs", "Herman Kamper" ]
https://www.isca-archive.org/interspeech_2021/jacobs21_interspeech.html
https://www.isca-archive.org/interspeech_2021/jacobs21_interspeech.pdf
10.21437/Interspeech.2021-461
1549-1553
@inproceedings{jacobs21_interspeech, title = {{Multilingual Transfer of Acoustic Word Embeddings Improves When Training on Languages Related to the Target Zero-Resource Language}}, author = {Christiaan Jacobs and Herman Kamper}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1549--1...
Acoustic word embedding models map variable duration speech segments to fixed dimensional vectors, enabling efficient speech search and discovery. Previous work explored how embeddings can be obtained in zero-resource settings where no labelled data is available in the target language. The current best approach uses tr...
2106.12834
title_snapshot
niekerk21_interspeech
Analyzing Speaker Information in Self-Supervised Models to Improve Zero-Resource Speech Processing
[ "Benjamin van Niekerk", "Leanne Nortje", "Matthew Baas", "Herman Kamper" ]
https://www.isca-archive.org/interspeech_2021/niekerk21_interspeech.html
https://www.isca-archive.org/interspeech_2021/niekerk21_interspeech.pdf
10.21437/Interspeech.2021-1182
1554-1558
@inproceedings{niekerk21_interspeech, title = {{Analyzing Speaker Information in Self-Supervised Models to Improve Zero-Resource Speech Processing}}, author = {Benjamin van Niekerk and Leanne Nortje and Matthew Baas and Herman Kamper}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {...
Contrastive predictive coding (CPC) aims to learn representations of speech by distinguishing future observations from a set of negative examples. Previous work has shown that linear classifiers trained on CPC features can accurately predict speaker and phone labels. However, it is unclear how the features actually cap...
2108.00917
title_snapshot
takahashi21_interspeech
Unsupervised Neural-Based Graph Clustering for Variable-Length Speech Representation Discovery of Zero-Resource Languages
[ "Shun Takahashi", "Sakriani Sakti", "Satoshi Nakamura" ]
https://www.isca-archive.org/interspeech_2021/takahashi21_interspeech.html
https://www.isca-archive.org/interspeech_2021/takahashi21_interspeech.pdf
10.21437/Interspeech.2021-1340
1559-1563
@inproceedings{takahashi21_interspeech, title = {{Unsupervised Neural-Based Graph Clustering for Variable-Length Speech Representation Discovery of Zero-Resource Languages}}, author = {Shun Takahashi and Sakriani Sakti and Satoshi Nakamura}, year = {2021}, booktitle = {{Interspeech 2021}}, pages ...
Discovering symbolic units from unannotated speech data is fundamental in zero resource speech technology. Previous studies focused on learning fixed-length frame units based on acoustic features. Although they achieve high quality, they also suffer from a high bit-rate due to time-frame encoding. In this work, to disc...
null
null
maekaku21_interspeech
Speech Representation Learning Combining Conformer CPC with Deep Cluster for the ZeroSpeech Challenge 2021
[ "Takashi Maekaku", "Xuankai Chang", "Yuya Fujita", "Li-Wei Chen", "Shinji Watanabe", "Alexander Rudnicky" ]
https://www.isca-archive.org/interspeech_2021/maekaku21_interspeech.html
https://www.isca-archive.org/interspeech_2021/maekaku21_interspeech.pdf
10.21437/Interspeech.2021-1503
1564-1568
@inproceedings{maekaku21_interspeech, title = {{Speech Representation Learning Combining Conformer CPC with Deep Cluster for the ZeroSpeech Challenge 2021}}, author = {Takashi Maekaku and Xuankai Chang and Yuya Fujita and Li-Wei Chen and Shinji Watanabe and Alexander Rudnicky}, year = {2021}, bookti...
We present a system for the Zero Resource Speech Challenge 2021, which combines a Contrastive Predictive Coding (CPC) with deep cluster. In deep cluster, we first prepare pseudo-labels obtained by clustering the outputs of a CPC network with k-means. Then, we train an additional autoregressive model to classify the pre...
2107.05899
title_snapshot
cui21_interspeech
Identifying Indicators of Vulnerability from Short Speech Segments Using Acoustic and Textual Features
[ "Xia Cui", "Amila Gamage", "Terry Hanley", "Tingting Mu" ]
https://www.isca-archive.org/interspeech_2021/cui21_interspeech.html
https://www.isca-archive.org/interspeech_2021/cui21_interspeech.pdf
10.21437/Interspeech.2021-1525
1569-1573
@inproceedings{cui21_interspeech, title = {{Identifying Indicators of Vulnerability from Short Speech Segments Using Acoustic and Textual Features}}, author = {Xia Cui and Amila Gamage and Terry Hanley and Tingting Mu}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1569--1573}, d...
In order to protect vulnerable people in telemarketing, organisations have to investigate the speech recordings to identify them first. Typically, the investigation is manually conducted. As such, the procedure is costly and time-consuming. With an automatic vulnerability detection system, more vulnerable people can be...
null
null
dunbar21_interspeech
The Zero Resource Speech Challenge 2021: Spoken Language Modelling
[ "Ewan Dunbar", "Mathieu Bernard", "Nicolas Hamilakis", "Tu Anh Nguyen", "Maureen de Seyssel", "Patricia Rozé", "Morgane Rivière", "Eugene Kharitonov", "Emmanuel Dupoux" ]
https://www.isca-archive.org/interspeech_2021/dunbar21_interspeech.html
https://www.isca-archive.org/interspeech_2021/dunbar21_interspeech.pdf
10.21437/Interspeech.2021-1755
1574-1578
@inproceedings{dunbar21_interspeech, title = {{The Zero Resource Speech Challenge 2021: Spoken Language Modelling}}, author = {Ewan Dunbar and Mathieu Bernard and Nicolas Hamilakis and Tu Anh Nguyen and Maureen de Seyssel and Patricia Rozé and Morgane Rivière and Eugene Kharitonov and Emmanuel Dupoux}, yea...
We present the Zero Resource Speech Challenge 2021, which asks participants to learn a language model directly from audio, without any text or labels. The challenge is based on the Libri-light dataset, which provides up to 60k hours of audio from English audio books without any associated text. We provide a pipeline ba...
2104.14700
title_snapshot
gudur21_interspeech
Zero-Shot Federated Learning with New Classes for Audio Classification
[ "Gautham Krishna Gudur", "Satheesh Kumar Perepu" ]
https://www.isca-archive.org/interspeech_2021/gudur21_interspeech.html
https://www.isca-archive.org/interspeech_2021/gudur21_interspeech.pdf
10.21437/Interspeech.2021-2264
1579-1583
@inproceedings{gudur21_interspeech, title = {{Zero-Shot Federated Learning with New Classes for Audio Classification}}, author = {Gautham Krishna Gudur and Satheesh Kumar Perepu}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1579--1583}, doi = {10.21437/Interspeech.2021-22...
Federated learning is an effective way of extracting insights from different user devices while preserving the privacy of users. However, new classes with completely unseen data distributions can stream across any device in a federated learning setting, whose data cannot be accessed by the global server or other users....
2106.10019
title_snapshot
rouditchenko21_interspeech
AVLnet: Learning Audio-Visual Language Representations from Instructional Videos
[ "Andrew Rouditchenko", "Angie Boggust", "David Harwath", "Brian Chen", "Dhiraj Joshi", "Samuel Thomas", "Kartik Audhkhasi", "Hilde Kuehne", "Rameswar Panda", "Rogerio Feris", "Brian Kingsbury", "Michael Picheny", "Antonio Torralba", "James Glass" ]
https://www.isca-archive.org/interspeech_2021/rouditchenko21_interspeech.html
https://www.isca-archive.org/interspeech_2021/rouditchenko21_interspeech.pdf
10.21437/Interspeech.2021-1312
1584-1588
@inproceedings{rouditchenko21_interspeech, title = {{AVLnet: Learning Audio-Visual Language Representations from Instructional Videos}}, author = {Andrew Rouditchenko and Angie Boggust and David Harwath and Brian Chen and Dhiraj Joshi and Samuel Thomas and Kartik Audhkhasi and Hilde Kuehne and Rameswar Panda...
Current methods for learning visually grounded language from videos often rely on text annotation, such as human generated captions or machine generated automatic speech recognition (ASR) transcripts. In this work, we introduce the Audio-Video Language Network (AVLnet), a self-supervised network that learns a shared au...
2006.09199
title_snapshot
lee21b_interspeech
N-Singer: A Non-Autoregressive Korean Singing Voice Synthesis System for Pronunciation Enhancement
[ "Gyeong-Hoon Lee", "Tae-Woo Kim", "Hanbin Bae", "Min-Ji Lee", "Young-Ik Kim", "Hoon-Young Cho" ]
https://www.isca-archive.org/interspeech_2021/lee21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/lee21b_interspeech.pdf
10.21437/Interspeech.2021-239
1589-1593
@inproceedings{lee21b_interspeech, title = {{N-Singer: A Non-Autoregressive Korean Singing Voice Synthesis System for Pronunciation Enhancement}}, author = {Gyeong-Hoon Lee and Tae-Woo Kim and Hanbin Bae and Min-Ji Lee and Young-Ik Kim and Hoon-Young Cho}, year = {2021}, booktitle = {{Interspeech 20...
Recently, end-to-end Korean singing voice systems have been designed to generate realistic singing voices. However, these systems still suffer from a lack of robustness in terms of pronunciation accuracy. In this paper, we propose N-Singer, a non-autoregressive Korean singing voice system, to synthesize accurate and pr...
2106.15205
title_snapshot
maniati21_interspeech
Cross-Lingual Low Resource Speaker Adaptation Using Phonological Features
[ "Georgia Maniati", "Nikolaos Ellinas", "Konstantinos Markopoulos", "Georgios Vamvoukakis", "June Sig Sung", "Hyoungmin Park", "Aimilios Chalamandaris", "Pirros Tsiakoulis" ]
https://www.isca-archive.org/interspeech_2021/maniati21_interspeech.html
https://www.isca-archive.org/interspeech_2021/maniati21_interspeech.pdf
10.21437/Interspeech.2021-327
1594-1598
@inproceedings{maniati21_interspeech, title = {{Cross-Lingual Low Resource Speaker Adaptation Using Phonological Features}}, author = {Georgia Maniati and Nikolaos Ellinas and Konstantinos Markopoulos and Georgios Vamvoukakis and June Sig Sung and Hyoungmin Park and Aimilios Chalamandaris and Pirros Tsiakoul...
The idea of using phonological features instead of phonemes as input to sequence-to-sequence TTS has been recently proposed for zero-shot multilingual speech synthesis. This approach is useful for code-switching, as it facilitates the seamless uttering of foreign text embedded in a stream of native text. In our work, w...
2111.09075
title_snapshot
zhan21_interspeech
Improve Cross-Lingual Text-To-Speech Synthesis on Monolingual Corpora with Pitch Contour Information
[ "Haoyue Zhan", "Haitong Zhang", "Wenjie Ou", "Yue Lin" ]
https://www.isca-archive.org/interspeech_2021/zhan21_interspeech.html
https://www.isca-archive.org/interspeech_2021/zhan21_interspeech.pdf
10.21437/Interspeech.2021-474
1599-1603
@inproceedings{zhan21_interspeech, title = {{Improve Cross-Lingual Text-To-Speech Synthesis on Monolingual Corpora with Pitch Contour Information}}, author = {Haoyue Zhan and Haitong Zhang and Wenjie Ou and Yue Lin}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1599--1603}, doi ...
Cross-lingual text-to-speech (TTS) synthesis on monolingual corpora is still a challenging task, especially when many kinds of languages are involved. In this paper, we improve the cross-lingual TTS model on monolingual corpora with pitch contour information. We propose a method to obtain pitch contour sequences for di...
null
null
yang21d_interspeech
Cross-Lingual Voice Conversion with Disentangled Universal Linguistic Representations
[ "Zhenchuan Yang", "Weibin Zhang", "Yufei Liu", "Xiaofen Xing" ]
https://www.isca-archive.org/interspeech_2021/yang21d_interspeech.html
https://www.isca-archive.org/interspeech_2021/yang21d_interspeech.pdf
10.21437/Interspeech.2021-552
1604-1608
@inproceedings{yang21d_interspeech, title = {{Cross-Lingual Voice Conversion with Disentangled Universal Linguistic Representations}}, author = {Zhenchuan Yang and Weibin Zhang and Yufei Liu and Xiaofen Xing}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1604--1608}, doi =...
Intra-lingual voice conversion has achieved great progress recently in terms of naturalness and similarity. However, in cross-lingual voice conversion, there is still an urgent need to improve the quality of the converted speech, especially with nonparallel training data. Previous works usually use Phonetic Posteriorgr...
null
null
liu21e_interspeech
EfficientSing: A Chinese Singing Voice Synthesis System Using Duration-Free Acoustic Model and HiFi-GAN Vocoder
[ "Zhengchen Liu", "Chenfeng Miao", "Qingying Zhu", "Minchuan Chen", "Jun Ma", "Shaojun Wang", "Jing Xiao" ]
https://www.isca-archive.org/interspeech_2021/liu21e_interspeech.html
https://www.isca-archive.org/interspeech_2021/liu21e_interspeech.pdf
10.21437/Interspeech.2021-771
1609-1613
@inproceedings{liu21e_interspeech, title = {{EfficientSing: A Chinese Singing Voice Synthesis System Using Duration-Free Acoustic Model and HiFi-GAN Vocoder}}, author = {Zhengchen Liu and Chenfeng Miao and Qingying Zhu and Minchuan Chen and Jun Ma and Shaojun Wang and Jing Xiao}, year = {2021}, book...
In this paper, we present EfficientSing, a Chinese singing voice synthesis (SVS) system based on a non-autoregressive duration-free acoustic model and HiFi-GAN neural vocoder. Different from many existing SVS methods, no auxiliary duration prediction module is needed in this work, since a newly proposed monotonic align...
null
null
xin21_interspeech
Cross-Lingual Speaker Adaptation Using Domain Adaptation and Speaker Consistency Loss for Text-To-Speech Synthesis
[ "Detai Xin", "Yuki Saito", "Shinnosuke Takamichi", "Tomoki Koriyama", "Hiroshi Saruwatari" ]
https://www.isca-archive.org/interspeech_2021/xin21_interspeech.html
https://www.isca-archive.org/interspeech_2021/xin21_interspeech.pdf
10.21437/Interspeech.2021-897
1614-1618
@inproceedings{xin21_interspeech, title = {{Cross-Lingual Speaker Adaptation Using Domain Adaptation and Speaker Consistency Loss for Text-To-Speech Synthesis}}, author = {Detai Xin and Yuki Saito and Shinnosuke Takamichi and Tomoki Koriyama and Hiroshi Saruwatari}, year = {2021}, booktitle = {{Inte...
We present a cross-lingual speaker adaptation method based on domain adaptation and a speaker consistency loss for text-to-speech (TTS) synthesis. Existing monolingual speaker adaptation methods based on direct fine-tuning are not applicable for cross-lingual data. The proposed method first trains a language-independen...
null
null
shang21_interspeech
Incorporating Cross-Speaker Style Transfer for Multi-Language Text-to-Speech
[ "Zengqiang Shang", "Zhihua Huang", "Haozhe Zhang", "Pengyuan Zhang", "Yonghong Yan" ]
https://www.isca-archive.org/interspeech_2021/shang21_interspeech.html
https://www.isca-archive.org/interspeech_2021/shang21_interspeech.pdf
10.21437/Interspeech.2021-1265
1619-1623
@inproceedings{shang21_interspeech, title = {{Incorporating Cross-Speaker Style Transfer for Multi-Language Text-to-Speech}}, author = {Zengqiang Shang and Zhihua Huang and Haozhe Zhang and Pengyuan Zhang and Yonghong Yan}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1619--1623},...
Recently multilingual TTS systems using only monolingual datasets have obtained significant improvement. However, the quality of cross-language speech synthesis is not comparable to the speaker’s own language and often comes with a heavy foreign accent. This paper proposed a multi-speaker multi-style multi-language spe...
null
null
kesim21_interspeech
Investigating Contributions of Speech and Facial Landmarks for Talking Head Generation
[ "Ege Kesim", "Engin Erzin" ]
https://www.isca-archive.org/interspeech_2021/kesim21_interspeech.html
https://www.isca-archive.org/interspeech_2021/kesim21_interspeech.pdf
10.21437/Interspeech.2021-1585
1624-1628
@inproceedings{kesim21_interspeech, title = {{Investigating Contributions of Speech and Facial Landmarks for Talking Head Generation}}, author = {Ege Kesim and Engin Erzin}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1624--1628}, doi = {10.21437/Interspeech.2021-1585}, ...
Talking head generation is an active research problem. It has been widely studied as a direct speech-to-video or two stage speech-to-landmarks-to-video mapping problem. In this study, our main motivation is to assess individual and joint contributions of the speech and facial landmarks to the talking head generation qu...
null
null
si21b_interspeech
Speech2Video: Cross-Modal Distillation for Speech to Video Generation
[ "Shijing Si", "Jianzong Wang", "Xiaoyang Qu", "Ning Cheng", "Wenqi Wei", "Xinghua Zhu", "Jing Xiao" ]
https://www.isca-archive.org/interspeech_2021/si21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/si21b_interspeech.pdf
10.21437/Interspeech.2021-1996
1629-1633
@inproceedings{si21b_interspeech, title = {{Speech2Video: Cross-Modal Distillation for Speech to Video Generation}}, author = {Shijing Si and Jianzong Wang and Xiaoyang Qu and Ning Cheng and Wenqi Wei and Xinghua Zhu and Jing Xiao}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {162...
This paper investigates a novel task of talking face video generation solely from speeches. The speech-to-video generation technique can spark interesting applications in entertainment, customer service, and human-computer-interaction industries. Indeed, the timbre, accent and speed in speeches could contain rich infor...
2107.04806
title_snapshot
lee21c_interspeech
NU-Wave: A Diffusion Probabilistic Model for Neural Audio Upsampling
[ "Junhyeok Lee", "Seungu Han" ]
https://www.isca-archive.org/interspeech_2021/lee21c_interspeech.html
https://www.isca-archive.org/interspeech_2021/lee21c_interspeech.pdf
10.21437/Interspeech.2021-36
1634-1638
@inproceedings{lee21c_interspeech, title = {{NU-Wave: A Diffusion Probabilistic Model for Neural Audio Upsampling}}, author = {Junhyeok Lee and Seungu Han}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1634--1638}, doi = {10.21437/Interspeech.2021-36}, issn = {2958-...
In this work, we introduce NU-Wave , the first neural audio upsampling model to produce waveforms of sampling rate 48kHz from coarse 16kHz or 24kHz inputs, while prior works could generate only up to 16kHz. NU-Wave is the first diffusion probabilistic model for audio super-resolution which is engineered based on neural...
2104.02321
title_snapshot
lin21c_interspeech
QISTA-Net-Audio: Audio Super-Resolution via Non-Convex ℓ_q-Norm Minimization
[ "Gang-Xuan Lin", "Shih-Wei Hu", "Yen-Ju Lu", "Yu Tsao", "Chun-Shien Lu" ]
https://www.isca-archive.org/interspeech_2021/lin21c_interspeech.html
https://www.isca-archive.org/interspeech_2021/lin21c_interspeech.pdf
10.21437/Interspeech.2021-670
1639-1643
@inproceedings{lin21c_interspeech, title = {{QISTA-Net-Audio: Audio Super-Resolution via Non-Convex ℓ_q-Norm Minimization}}, author = {Gang-Xuan Lin and Shih-Wei Hu and Yen-Ju Lu and Yu Tsao and Chun-Shien Lu}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1639--1643}, doi ...
Audio super-resolution (ASR) aims to reconstruct the high-resolution signal from its corresponding low-resolution one, which is hard while the correlation between them is low. In this paper, we propose a learning model, QISTA-Net-Audio, to solve ASR in a paradigm of linear inverse problem. QISTA-Net-Audio is composed o...
null
null
wen21_interspeech
X-net: A Joint Scale Down and Scale Up Method for Voice Call
[ "Liang Wen", "Lizhong Wang", "Xue Wen", "Yuxing Zheng", "Youngo Park", "Kwang Pyo Choi" ]
https://www.isca-archive.org/interspeech_2021/wen21_interspeech.html
https://www.isca-archive.org/interspeech_2021/wen21_interspeech.pdf
10.21437/Interspeech.2021-812
1644-1648
@inproceedings{wen21_interspeech, title = {{X-net: A Joint Scale Down and Scale Up Method for Voice Call}}, author = {Liang Wen and Lizhong Wang and Xue Wen and Yuxing Zheng and Youngo Park and Kwang Pyo Choi}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1644--1648}, doi ...
This paper proposes X-net, a jointly learned scale-down and scale-up architecture for data pre- and post-processing in voice calls, as a means to bandwidth extension over band-limited channels. Scale-down and scale-up are deployed separately on transmitter and receiver to perform down- and upsampling. Separate supervis...
null
null
zhang21k_interspeech
WSRGlow: A Glow-Based Waveform Generative Model for Audio Super-Resolution
[ "Kexun Zhang", "Yi Ren", "Changliang Xu", "Zhou Zhao" ]
https://www.isca-archive.org/interspeech_2021/zhang21k_interspeech.html
https://www.isca-archive.org/interspeech_2021/zhang21k_interspeech.pdf
10.21437/Interspeech.2021-892
1649-1653
@inproceedings{zhang21k_interspeech, title = {{WSRGlow: A Glow-Based Waveform Generative Model for Audio Super-Resolution}}, author = {Kexun Zhang and Yi Ren and Changliang Xu and Zhou Zhao}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1649--1653}, doi = {10.21437/Intersp...
Audio super-resolution is the task of constructing a high-resolution (HR) audio from a low-resolution (LR) audio by adding the missing band. Previous methods based on convolutional neural networks and mean squared error training objective have relatively low performance, while adversarial generative models are difficul...
2106.08507
title_snapshot
yi21_interspeech
Half-Truth: A Partially Fake Audio Detection Dataset
[ "Jiangyan Yi", "Ye Bai", "Jianhua Tao", "Haoxin Ma", "Zhengkun Tian", "Chenglong Wang", "Tao Wang", "Ruibo Fu" ]
https://www.isca-archive.org/interspeech_2021/yi21_interspeech.html
https://www.isca-archive.org/interspeech_2021/yi21_interspeech.pdf
10.21437/Interspeech.2021-930
1654-1658
@inproceedings{yi21_interspeech, title = {{Half-Truth: A Partially Fake Audio Detection Dataset}}, author = {Jiangyan Yi and Ye Bai and Jianhua Tao and Haoxin Ma and Zhengkun Tian and Chenglong Wang and Tao Wang and Ruibo Fu}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1654--165...
Diverse promising datasets have been designed to further the development of fake audio detection, such as ASVspoof databases. However, previous datasets ignore an attacking situation, in which the hacker hides some small fake clips in real speech audio. This poses a serious threat since that it is difficult to distingu...
2104.03617
title_snapshot
chettri21_interspeech
Data Quality as Predictor of Voice Anti-Spoofing Generalization
[ "Bhusan Chettri", "Rosa González Hautamäki", "Md. Sahidullah", "Tomi Kinnunen" ]
https://www.isca-archive.org/interspeech_2021/chettri21_interspeech.html
https://www.isca-archive.org/interspeech_2021/chettri21_interspeech.pdf
10.21437/Interspeech.2021-1180
1659-1663
@inproceedings{chettri21_interspeech, title = {{Data Quality as Predictor of Voice Anti-Spoofing Generalization}}, author = {Bhusan Chettri and Rosa González Hautamäki and Md. Sahidullah and Tomi Kinnunen}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1659--1663}, doi = {1...
Voice anti-spoofing aims at classifying a given utterance either as a bonafide human sample, or a spoofing attack (e.g. synthetic or replayed sample). Many anti-spoofing methods have been proposed but most of them fail to generalize across domains (corpora) — and we do not know why . We outline a novel interpretative f...
2103.14602
title_snapshot
cheon21_interspeech
Coded Speech Enhancement Using Neural Network-Based Vector-Quantized Residual Features
[ "Youngju Cheon", "Soojoong Hwang", "Sangwook Han", "Inseon Jang", "Jong Won Shin" ]
https://www.isca-archive.org/interspeech_2021/cheon21_interspeech.html
https://www.isca-archive.org/interspeech_2021/cheon21_interspeech.pdf
10.21437/Interspeech.2021-1204
1664-1668
@inproceedings{cheon21_interspeech, title = {{Coded Speech Enhancement Using Neural Network-Based Vector-Quantized Residual Features}}, author = {Youngju Cheon and Soojoong Hwang and Sangwook Han and Inseon Jang and Jong Won Shin}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1664...
Various approaches have been proposed to improve the quality of the speech coded at low bitrates. Recently, deep neural networks have also been used for speech coding, providing a high quality of speech with low bitrates. Although designing an entire codec with neural networks may be more effective, backward compatibil...
null
null
drude21_interspeech
Multi-Channel Opus Compression for Far-Field Automatic Speech Recognition with a Fixed Bitrate Budget
[ "Lukas Drude", "Jahn Heymann", "Andreas Schwarz", "Jean-Marc Valin" ]
https://www.isca-archive.org/interspeech_2021/drude21_interspeech.html
https://www.isca-archive.org/interspeech_2021/drude21_interspeech.pdf
10.21437/Interspeech.2021-1214
1669-1673
@inproceedings{drude21_interspeech, title = {{Multi-Channel Opus Compression for Far-Field Automatic Speech Recognition with a Fixed Bitrate Budget}}, author = {Lukas Drude and Jahn Heymann and Andreas Schwarz and Jean-Marc Valin}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1669...
Automatic speech recognition (ASR) in the cloud allows the use of larger models and more powerful multi-channel signal processing front-ends compared to on-device processing. However, it also adds an inherent latency due to the transmission of the audio signal, especially when transmitting multiple channels of a microp...
2106.07994
title_snapshot
siegert21_interspeech
Effects of Prosodic Variations on Accidental Triggers of a Commercial Voice Assistant
[ "Ingo Siegert" ]
https://www.isca-archive.org/interspeech_2021/siegert21_interspeech.html
https://www.isca-archive.org/interspeech_2021/siegert21_interspeech.pdf
10.21437/Interspeech.2021-1354
1674-1678
@inproceedings{siegert21_interspeech, title = {{Effects of Prosodic Variations on Accidental Triggers of a Commercial Voice Assistant}}, author = {Ingo Siegert}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1674--1678}, doi = {10.21437/Interspeech.2021-1354}, issn =...
The use of modern voice assistants has rapidly grown and they can be found in more and more households. By design, these systems have to scan every sound in their surroundings waiting for their respective wake-word before being able to react to the users’ commands. The drawback of this method is that phonetic similar e...
null
null
gabrys21_interspeech
Improving the Expressiveness of Neural Vocoding with Non-Affine Normalizing Flows
[ "Adam Gabryś", "Yunlong Jiao", "Viacheslav Klimkov", "Daniel Korzekwa", "Roberto Barra-Chicote" ]
https://www.isca-archive.org/interspeech_2021/gabrys21_interspeech.html
https://www.isca-archive.org/interspeech_2021/gabrys21_interspeech.pdf
10.21437/Interspeech.2021-1555
1679-1683
@inproceedings{gabrys21_interspeech, title = {{Improving the Expressiveness of Neural Vocoding with Non-Affine Normalizing Flows}}, author = {Adam Gabryś and Yunlong Jiao and Viacheslav Klimkov and Daniel Korzekwa and Roberto Barra-Chicote}, year = {2021}, booktitle = {{Interspeech 2021}}, pages ...
This paper proposes a general enhancement to the Normalizing Flows (NF) used in neural vocoding. As a case study, we improve expressive speech vocoding with a revamped Parallel Wavenet (PW). Specifically, we propose to extend the affine transformation of PW to the more expressive invertible non-affine function. The gre...
2106.08649
title_snapshot
prajapati21_interspeech
Voice Privacy Through x-Vector and CycleGAN-Based Anonymization
[ "Gauri P. Prajapati", "Dipesh K. Singh", "Preet P. Amin", "Hemant A. Patil" ]
https://www.isca-archive.org/interspeech_2021/prajapati21_interspeech.html
https://www.isca-archive.org/interspeech_2021/prajapati21_interspeech.pdf
10.21437/Interspeech.2021-1573
1684-1688
@inproceedings{prajapati21_interspeech, title = {{Voice Privacy Through x-Vector and CycleGAN-Based Anonymization}}, author = {Gauri P. Prajapati and Dipesh K. Singh and Preet P. Amin and Hemant A. Patil}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1684--1688}, doi = {10...
With the rise in usage of voice assistants and spoken language interfaces, important concerns regarding voice data privacy have been prompted. In an attempt to reduce the threat of attacks on voice data, in this paper, we propose a speaker anonymization system based on CycleGAN. This method modifies the speaker’s gende...
null
null
lin21d_interspeech
A Two-Stage Approach to Speech Bandwidth Extension
[ "Ju Lin", "Yun Wang", "Kaustubh Kalgaonkar", "Gil Keren", "Didi Zhang", "Christian Fuegen" ]
https://www.isca-archive.org/interspeech_2021/lin21d_interspeech.html
https://www.isca-archive.org/interspeech_2021/lin21d_interspeech.pdf
10.21437/Interspeech.2021-1941
1689-1693
@inproceedings{lin21d_interspeech, title = {{A Two-Stage Approach to Speech Bandwidth Extension}}, author = {Ju Lin and Yun Wang and Kaustubh Kalgaonkar and Gil Keren and Didi Zhang and Christian Fuegen}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1689--1693}, doi = {10....
Algorithms for speech bandwidth extension (BWE) may work in either the time domain or the frequency domain. Time-domain methods often do not sufficiently recover the high-frequency content of speech signals; frequency-domain methods are better at recovering the spectral envelope, but have difficulty reconstructing the ...
null
null
byun21_interspeech
Development of a Psychoacoustic Loss Function for the Deep Neural Network (DNN)-Based Speech Coder
[ "Joon Byun", "Seungmin Shin", "Youngcheol Park", "Jongmo Sung", "Seungkwon Beack" ]
https://www.isca-archive.org/interspeech_2021/byun21_interspeech.html
https://www.isca-archive.org/interspeech_2021/byun21_interspeech.pdf
10.21437/Interspeech.2021-2151
1694-1698
@inproceedings{byun21_interspeech, title = {{Development of a Psychoacoustic Loss Function for the Deep Neural Network (DNN)-Based Speech Coder}}, author = {Joon Byun and Seungmin Shin and Youngcheol Park and Jongmo Sung and Seungkwon Beack}, year = {2021}, booktitle = {{Interspeech 2021}}, pages ...
This paper presents a loss function to compensate for the perceptual loss of the deep neural network (DNN)-based speech coder. By utilizing the psychoacoustic model (PAM), we design a loss function to maximize the mask-to-noise ratio (MNR) in multi-resolution Mel-frequency scales. Also, a perceptual entropy (PE)-based ...
null
null
stoidis21_interspeech
Protecting Gender and Identity with Disentangled Speech Representations
[ "Dimitrios Stoidis", "Andrea Cavallaro" ]
https://www.isca-archive.org/interspeech_2021/stoidis21_interspeech.html
https://www.isca-archive.org/interspeech_2021/stoidis21_interspeech.pdf
10.21437/Interspeech.2021-2163
1699-1703
@inproceedings{stoidis21_interspeech, title = {{Protecting Gender and Identity with Disentangled Speech Representations}}, author = {Dimitrios Stoidis and Andrea Cavallaro}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1699--1703}, doi = {10.21437/Interspeech.2021-2163}, ...
Besides its linguistic content, our speech is rich in biometric information that can be inferred by classifiers. Learning privacy-preserving representations for speech signals enables downstream tasks without sharing unnecessary, private information about an individual. In this paper, we show that protecting gender inf...
2104.11051
title_snapshot
aldholmi21_interspeech
Perception of Standard Arabic Synthetic Speech Rate
[ "Yahya Aldholmi", "Rawan Aldhafyan", "Asma Alqahtani" ]
https://www.isca-archive.org/interspeech_2021/aldholmi21_interspeech.html
https://www.isca-archive.org/interspeech_2021/aldholmi21_interspeech.pdf
10.21437/Interspeech.2021-39
1704-1707
@inproceedings{aldholmi21_interspeech, title = {{Perception of Standard Arabic Synthetic Speech Rate}}, author = {Yahya Aldholmi and Rawan Aldhafyan and Asma Alqahtani}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1704--1707}, doi = {10.21437/Interspeech.2021-39}, issn ...
This experiment investigated how Arabic speakers perceive synthetic Standard Arabic speech rate produced by Google TTS, at normal vs. accelerated rates. Twenty syntactically identical Standard Arabic sentences with a similar length ( M = 22 syllables per sentence, SD = 1) were auditorily presented in a female voice to ...
null
null
kishiyama21_interspeech
The Influence of Parallel Processing on Illusory Vowels
[ "Takeshi Kishiyama" ]
https://www.isca-archive.org/interspeech_2021/kishiyama21_interspeech.html
https://www.isca-archive.org/interspeech_2021/kishiyama21_interspeech.pdf
10.21437/Interspeech.2021-89
1708-1712
@inproceedings{kishiyama21_interspeech, title = {{The Influence of Parallel Processing on Illusory Vowels}}, author = {Takeshi Kishiyama}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1708--1712}, doi = {10.21437/Interspeech.2021-89}, issn = {2958-1796}, }
Research has shown that listeners perceive illusory vowels inside consonant clusters that are not allowed in their L1. This phenomenon has been examined using several psycholinguistic and computational models, including hidden Markov models (HMMs), applied to human phoneme perception. However, the inference algorithm o...
null
null
chingacham21_interspeech
Exploring the Potential of Lexical Paraphrases for Mitigating Noise-Induced Comprehension Errors
[ "Anupama Chingacham", "Vera Demberg", "Dietrich Klakow" ]
https://www.isca-archive.org/interspeech_2021/chingacham21_interspeech.html
https://www.isca-archive.org/interspeech_2021/chingacham21_interspeech.pdf
10.21437/Interspeech.2021-306
1713-1717
@inproceedings{chingacham21_interspeech, title = {{Exploring the Potential of Lexical Paraphrases for Mitigating Noise-Induced Comprehension Errors}}, author = {Anupama Chingacham and Vera Demberg and Dietrich Klakow}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1713--1717}, do...
Listening in noisy environments can be difficult even for individuals with a normal hearing thresholds. The speech signal can be masked by noise, which may lead to word misperceptions on the side of the listener, and overall difficulty to understand the message. To mitigate hearing difficulties on listeners, a co-opera...
2107.08337
title_snapshot
simantiraki21_interspeech
SpeechAdjuster: A Tool for Investigating Listener Preferences and Speech Intelligibility
[ "Olympia Simantiraki", "Martin Cooke" ]
https://www.isca-archive.org/interspeech_2021/simantiraki21_interspeech.html
https://www.isca-archive.org/interspeech_2021/simantiraki21_interspeech.pdf
10.21437/Interspeech.2021-324
1718-1722
@inproceedings{simantiraki21_interspeech, title = {{ SpeechAdjuster: A Tool for Investigating Listener Preferences and Speech Intelligibility}}, author = {Olympia Simantiraki and Martin Cooke}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1718--1722}, doi = {10.21437/Inter...
Most of what we know about speech perception has been gleaned from tests in which listeners respond to stimuli chosen by an experimenter. This paper presents SpeechAdjuster, an open source tool that reverses the roles of listener and experimenter by allowing listeners direct control of speech characteristics in real-t...
null
null
saito21_interspeech
VocalTurk: Exploring Feasibility of Crowdsourced Speaker Identification
[ "Susumu Saito", "Yuta Ide", "Teppei Nakano", "Tetsuji Ogawa" ]
https://www.isca-archive.org/interspeech_2021/saito21_interspeech.html
https://www.isca-archive.org/interspeech_2021/saito21_interspeech.pdf
10.21437/Interspeech.2021-464
1723-1727
@inproceedings{saito21_interspeech, title = {{VocalTurk: Exploring Feasibility of Crowdsourced Speaker Identification}}, author = {Susumu Saito and Yuta Ide and Teppei Nakano and Tetsuji Ogawa}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1723--1727}, doi = {10.21437/Inte...
This paper presents VocalTurk, a feasibility study of crowdsourced speaker identification based on our worker dataset collected in Amazon Mechanical Turk. Crowdsourced data labeling has already been acknowledged in speech data processing nowadays, but empirical analysis that answer to common questions such as “ how acc...
null
null
xu21e_interspeech
Effects of Aging and Age-Related Hearing Loss on Talker Discrimination
[ "Min Xu", "Jing Shao", "Lan Wang" ]
https://www.isca-archive.org/interspeech_2021/xu21e_interspeech.html
https://www.isca-archive.org/interspeech_2021/xu21e_interspeech.pdf
10.21437/Interspeech.2021-682
1728-1732
@inproceedings{xu21e_interspeech, title = {{Effects of Aging and Age-Related Hearing Loss on Talker Discrimination}}, author = {Min Xu and Jing Shao and Lan Wang}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1728--1732}, doi = {10.21437/Interspeech.2021-682}, issn ...
Paralinguistic information is as important as linguistic information. Being familiar with talker’s voice may facilitate speech perception, especially in challenging conditions. Previous studies have suggested that aging and age-related hearing loss lead to the deterioration of the phonetic and phonological processing a...
null
null
zhang21l_interspeech
Relationships Between Perceptual Distinctiveness, Articulatory Complexity and Functional Load in Speech Communication
[ "Yuqing Zhang", "Zhu Li", "Bin Wu", "Yanlu Xie", "Binghuai Lin", "Jinsong Zhang" ]
https://www.isca-archive.org/interspeech_2021/zhang21l_interspeech.html
https://www.isca-archive.org/interspeech_2021/zhang21l_interspeech.pdf
10.21437/Interspeech.2021-721
1733-1737
@inproceedings{zhang21l_interspeech, title = {{Relationships Between Perceptual Distinctiveness, Articulatory Complexity and Functional Load in Speech Communication}}, author = {Yuqing Zhang and Zhu Li and Bin Wu and Yanlu Xie and Binghuai Lin and Jinsong Zhang}, year = {2021}, booktitle = {{Intersp...
Work on communicative efficiency has hypothesized that phonological contrasts signaling more meaning distinctions (i.e., of high functional load (FL)) tend to have the least articulatory complexity and the highest perceptual salience. However, only a few studies have examined the preference for perceptual distinctivene...
null
null
terblanche21_interspeech
Human Spoofing Detection Performance on Degraded Speech
[ "Camryn Terblanche", "Philip Harrison", "Amelia J. Gully" ]
https://www.isca-archive.org/interspeech_2021/terblanche21_interspeech.html
https://www.isca-archive.org/interspeech_2021/terblanche21_interspeech.pdf
10.21437/Interspeech.2021-1225
1738-1742
@inproceedings{terblanche21_interspeech, title = {{Human Spoofing Detection Performance on Degraded Speech}}, author = {Camryn Terblanche and Philip Harrison and Amelia J. Gully}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1738--1742}, doi = {10.21437/Interspeech.2021-12...
Over the past few years attention has been focused on the automatic detection of spoofing in the context of automatic speaker verification (ASV) systems. However, little is known about how well humans perform at detecting spoofed speech, particularly under degraded conditions. Using the latest synthesis technologies fr...
null
null
einfeldt21_interspeech
Reliable Estimates of Interpretable Cue Effects with Active Learning in Psycholinguistic Research
[ "Marieke Einfeldt", "Rita Sevastjanova", "Katharina Zahner-Ritter", "Ekaterina Kazak", "Bettina Braun" ]
https://www.isca-archive.org/interspeech_2021/einfeldt21_interspeech.html
https://www.isca-archive.org/interspeech_2021/einfeldt21_interspeech.pdf
10.21437/Interspeech.2021-1524
1743-1747
@inproceedings{einfeldt21_interspeech, title = {{Reliable Estimates of Interpretable Cue Effects with Active Learning in Psycholinguistic Research}}, author = {Marieke Einfeldt and Rita Sevastjanova and Katharina Zahner-Ritter and Ekaterina Kazak and Bettina Braun}, year = {2021}, booktitle = {{Inte...
Studying the relative weighting of different cues for the interpretation of a linguistic phenomenon is a core element in psycholinguistic research. This research needs to strike a balance between two things: generalisability to diverse lexical settings, which requires a high number of different lexicalisations and the ...
null
null
kumar21d_interspeech
Towards the Explainability of Multimodal Speech Emotion Recognition
[ "Puneet Kumar", "Vishesh Kaushik", "Balasubramanian Raman" ]
https://www.isca-archive.org/interspeech_2021/kumar21d_interspeech.html
https://www.isca-archive.org/interspeech_2021/kumar21d_interspeech.pdf
10.21437/Interspeech.2021-1718
1748-1752
@inproceedings{kumar21d_interspeech, title = {{Towards the Explainability of Multimodal Speech Emotion Recognition}}, author = {Puneet Kumar and Vishesh Kaushik and Balasubramanian Raman}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1748--1752}, doi = {10.21437/Interspeec...
In this paper, a multimodal speech emotion recognition system has been developed, and a novel technique to explain its predictions has been proposed. The audio and textual features are extracted separately using attention-based Gated Recurrent Unit (GRU) and pre-trained Bidirectional Encoder Representations from Transf...
null
null
zeng21_interspeech
Primacy of Mouth over Eyes: Eye Movement Evidence from Audiovisual Mandarin Lexical Tones and Vowels
[ "Biao Zeng", "Rui Wang", "Guoxing Yu", "Christian Dobel" ]
https://www.isca-archive.org/interspeech_2021/zeng21_interspeech.html
https://www.isca-archive.org/interspeech_2021/zeng21_interspeech.pdf
10.21437/Interspeech.2021-1741
1753-1756
@inproceedings{zeng21_interspeech, title = {{Primacy of Mouth over Eyes: Eye Movement Evidence from Audiovisual Mandarin Lexical Tones and Vowels}}, author = {Biao Zeng and Rui Wang and Guoxing Yu and Christian Dobel}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1753--1756}, do...
This study investigated Chinese speakers’ eye movements when they were asked to identify audiovisual Mandarin lexical tones and vowels. In the lexical tone identification task, Chinese speakers were presented with an audiovisual clip of Mandarin monosyllables (/ă/, /à/, /ĭ/, /ì/) and asked to identify whether the sylla...
null
null
ashihara21_interspeech
Investigating the Impact of Spectral and Temporal Degradation on End-to-End Automatic Speech Recognition Performance
[ "Takanori Ashihara", "Takafumi Moriya", "Makio Kashino" ]
https://www.isca-archive.org/interspeech_2021/ashihara21_interspeech.html
https://www.isca-archive.org/interspeech_2021/ashihara21_interspeech.pdf
10.21437/Interspeech.2021-2091
1757-1761
@inproceedings{ashihara21_interspeech, title = {{Investigating the Impact of Spectral and Temporal Degradation on End-to-End Automatic Speech Recognition Performance}}, author = {Takanori Ashihara and Takafumi Moriya and Makio Kashino}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = ...
Humans have a sophisticated capability to robustly handle incomplete sensory input, as often happens in real environments. In earlier studies, the robustness of human speech perception was observed qualitatively by spectrally and temporally degraded stimuli. The current study investigates how machine speech recognition...
null
null
nguyen21c_interspeech
Super-Human Performance in Online Low-Latency Recognition of Conversational Speech
[ "Thai-Son Nguyen", "Sebastian Stüker", "Alex Waibel" ]
https://www.isca-archive.org/interspeech_2021/nguyen21c_interspeech.html
https://www.isca-archive.org/interspeech_2021/nguyen21c_interspeech.pdf
10.21437/Interspeech.2021-1114
1762-1766
@inproceedings{nguyen21c_interspeech, title = {{Super-Human Performance in Online Low-Latency Recognition of Conversational Speech}}, author = {Thai-Son Nguyen and Sebastian Stüker and Alex Waibel}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1762--1766}, doi = {10.21437/...
Achieving super-human performance in recognizing human speech has been a goal for several decades as researchers have worked on increasingly challenging tasks. In the 1990’s it was discovered, that conversational speech between two humans turns out to be considerably more difficult than read speech as hesitations, disf...
2010.03449
title_snapshot
joshi21_interspeech
Multiple Softmax Architecture for Streaming Multilingual End-to-End ASR Systems
[ "Vikas Joshi", "Amit Das", "Eric Sun", "Rupesh R. Mehta", "Jinyu Li", "Yifan Gong" ]
https://www.isca-archive.org/interspeech_2021/joshi21_interspeech.html
https://www.isca-archive.org/interspeech_2021/joshi21_interspeech.pdf
10.21437/Interspeech.2021-1298
1767-1771
@inproceedings{joshi21_interspeech, title = {{Multiple Softmax Architecture for Streaming Multilingual End-to-End ASR Systems}}, author = {Vikas Joshi and Amit Das and Eric Sun and Rupesh R. Mehta and Jinyu Li and Yifan Gong}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1767--177...
Improving multilingual end-to-end (E2E) automatic speech recognition (ASR) systems have manifold advantages. They simplify the training strategy, are easier to scale and exhibit better performance over monolingual models. However, it is still challenging to use a single multilingual model to recognize multiple language...
null
null
le21_interspeech
Contextualized Streaming End-to-End Speech Recognition with Trie-Based Deep Biasing and Shallow Fusion
[ "Duc Le", "Mahaveer Jain", "Gil Keren", "Suyoun Kim", "Yangyang Shi", "Jay Mahadeokar", "Julian Chan", "Yuan Shangguan", "Christian Fuegen", "Ozlem Kalinli", "Yatharth Saraf", "Michael L. Seltzer" ]
https://www.isca-archive.org/interspeech_2021/le21_interspeech.html
https://www.isca-archive.org/interspeech_2021/le21_interspeech.pdf
10.21437/Interspeech.2021-1566
1772-1776
@inproceedings{le21_interspeech, title = {{Contextualized Streaming End-to-End Speech Recognition with Trie-Based Deep Biasing and Shallow Fusion}}, author = {Duc Le and Mahaveer Jain and Gil Keren and Suyoun Kim and Yangyang Shi and Jay Mahadeokar and Julian Chan and Yuan Shangguan and Christian Fuegen and ...
How to leverage dynamic contextual information in end-to-end speech recognition has remained an active research area. Previous solutions to this problem were either designed for specialized use cases that did not generalize well to open-domain scenarios, did not scale to large biasing lists, or underperformed on rare l...
2104.02194
title_snapshot
sainath21_interspeech
An Efficient Streaming Non-Recurrent On-Device End-to-End Model with Improvements to Rare-Word Modeling
[ "Tara N. Sainath", "Yanzhang He", "Arun Narayanan", "Rami Botros", "Ruoming Pang", "David Rybach", "Cyril Allauzen", "Ehsan Variani", "James Qin", "Quoc-Nam Le-The", "Shuo-Yiin Chang", "Bo Li", "Anmol Gulati", "Jiahui Yu", "Chung-Cheng Chiu", "Diamantino Caseiro", "Wei Li", "Qiao L...
https://www.isca-archive.org/interspeech_2021/sainath21_interspeech.html
https://www.isca-archive.org/interspeech_2021/sainath21_interspeech.pdf
10.21437/Interspeech.2021-206
1777-1781
@inproceedings{sainath21_interspeech, title = {{An Efficient Streaming Non-Recurrent On-Device End-to-End Model with Improvements to Rare-Word Modeling}}, author = {Tara N. Sainath and Yanzhang He and Arun Narayanan and Rami Botros and Ruoming Pang and David Rybach and Cyril Allauzen and Ehsan Variani and Ja...
On-device end-to-end (E2E) models have shown improvements over a conventional model on Search test sets in both quality, as measured by Word Error Rate (WER) [1], and latency [2], measured by the time the result is finalized after the user stops speaking. However, the E2E model is trained on a small fraction of audio-t...
null
null
lu21_interspeech
Streaming Multi-Talker Speech Recognition with Joint Speaker Identification
[ "Liang Lu", "Naoyuki Kanda", "Jinyu Li", "Yifan Gong" ]
https://www.isca-archive.org/interspeech_2021/lu21_interspeech.html
https://www.isca-archive.org/interspeech_2021/lu21_interspeech.pdf
10.21437/Interspeech.2021-207
1782-1786
@inproceedings{lu21_interspeech, title = {{Streaming Multi-Talker Speech Recognition with Joint Speaker Identification}}, author = {Liang Lu and Naoyuki Kanda and Jinyu Li and Yifan Gong}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1782--1786}, doi = {10.21437/Interspeec...
In multi-talker scenarios such as meetings and conversations, speech processing systems are usually required to transcribe the audio as well as identify the speakers for downstream applications. Since overlapped speech is common in this case, conventional approaches usually address this problem in a cascaded fashion th...
2104.02109
title_snapshot
moriya21_interspeech
Streaming End-to-End Speech Recognition for Hybrid RNN-T/Attention Architecture
[ "Takafumi Moriya", "Tomohiro Tanaka", "Takanori Ashihara", "Tsubasa Ochiai", "Hiroshi Sato", "Atsushi Ando", "Ryo Masumura", "Marc Delcroix", "Taichi Asami" ]
https://www.isca-archive.org/interspeech_2021/moriya21_interspeech.html
https://www.isca-archive.org/interspeech_2021/moriya21_interspeech.pdf
10.21437/Interspeech.2021-437
1787-1791
@inproceedings{moriya21_interspeech, title = {{Streaming End-to-End Speech Recognition for Hybrid RNN-T/Attention Architecture}}, author = {Takafumi Moriya and Tomohiro Tanaka and Takanori Ashihara and Tsubasa Ochiai and Hiroshi Sato and Atsushi Ando and Ryo Masumura and Marc Delcroix and Taichi Asami}, ye...
We present a novel architecture with its decoding approach for improving recurrent neural network-transducer (RNN-T) performance. RNN-T is promising for building time-synchronous automatic speech recognition (ASR) systems and thus enhancing streaming ASR applications. We note that encoder-decoder-based sequence-to-sequ...
null
null
schwarz21_interspeech
Improving RNN-T ASR Accuracy Using Context Audio
[ "Andreas Schwarz", "Ilya Sklyar", "Simon Wiesler" ]
https://www.isca-archive.org/interspeech_2021/schwarz21_interspeech.html
https://www.isca-archive.org/interspeech_2021/schwarz21_interspeech.pdf
10.21437/Interspeech.2021-542
1792-1796
@inproceedings{schwarz21_interspeech, title = {{Improving RNN-T ASR Accuracy Using Context Audio}}, author = {Andreas Schwarz and Ilya Sklyar and Simon Wiesler}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1792--1796}, doi = {10.21437/Interspeech.2021-542}, issn = ...
We present a training scheme for streaming automatic speech recognition (ASR) based on recurrent neural network transducers (RNN-T) which allows the encoder network to learn to exploit context audio from a stream, using segmented or partially labeled sequences of the stream during training. We show that the use of cont...
2011.10538
title_snapshot
huang21e_interspeech
HMM-Free Encoder Pre-Training for Streaming RNN Transducer
[ "Lu Huang", "Jingyu Sun", "Yufeng Tang", "Junfeng Hou", "Jinkun Chen", "Jun Zhang", "Zejun Ma" ]
https://www.isca-archive.org/interspeech_2021/huang21e_interspeech.html
https://www.isca-archive.org/interspeech_2021/huang21e_interspeech.pdf
10.21437/Interspeech.2021-586
1797-1801
@inproceedings{huang21e_interspeech, title = {{HMM-Free Encoder Pre-Training for Streaming RNN Transducer}}, author = {Lu Huang and Jingyu Sun and Yufeng Tang and Junfeng Hou and Jinkun Chen and Jun Zhang and Zejun Ma}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1797--1801}, d...
This work describes an encoder pre-training procedure using frame-wise label to improve the training of streaming recurrent neural network transducer (RNN-T) model. Streaming RNN-T trained from scratch usually performs worse than non-streaming RNN-T. Although it is common to address this issue through pre-training comp...
2104.10764
title_snapshot
cui21b_interspeech
Reducing Exposure Bias in Training Recurrent Neural Network Transducers
[ "Xiaodong Cui", "Brian Kingsbury", "George Saon", "David Haws", "Zoltán Tüske" ]
https://www.isca-archive.org/interspeech_2021/cui21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/cui21b_interspeech.pdf
10.21437/Interspeech.2021-587
1802-1806
@inproceedings{cui21b_interspeech, title = {{Reducing Exposure Bias in Training Recurrent Neural Network Transducers}}, author = {Xiaodong Cui and Brian Kingsbury and George Saon and David Haws and Zoltán Tüske}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1802--1806}, doi ...
When recurrent neural network transducers (RNNTs) are trained using the typical maximum likelihood criterion, the prediction network is trained only on ground truth label sequences. This leads to a mismatch during inference, known as exposure bias, when the model must deal with label sequences containing errors. In thi...
2108.10803
title_snapshot
doutre21_interspeech
Bridging the Gap Between Streaming and Non-Streaming ASR Systems by Distilling Ensembles of CTC and RNN-T Models
[ "Thibault Doutre", "Wei Han", "Chung-Cheng Chiu", "Ruoming Pang", "Olivier Siohan", "Liangliang Cao" ]
https://www.isca-archive.org/interspeech_2021/doutre21_interspeech.html
https://www.isca-archive.org/interspeech_2021/doutre21_interspeech.pdf
10.21437/Interspeech.2021-637
1807-1811
@inproceedings{doutre21_interspeech, title = {{Bridging the Gap Between Streaming and Non-Streaming ASR Systems by Distilling Ensembles of CTC and RNN-T Models}}, author = {Thibault Doutre and Wei Han and Chung-Cheng Chiu and Ruoming Pang and Olivier Siohan and Liangliang Cao}, year = {2021}, bookti...
Streaming end-to-end automatic speech recognition (ASR) systems are widely used in everyday applications that require transcribing speech to text in real-time. Their minimal latency makes them suitable for such tasks. Unlike their non-streaming counterparts, streaming models are constrained to be causal with no future ...
2104.14346
title_judge
audhkhasi21_interspeech
Mixture Model Attention: Flexible Streaming and Non-Streaming Automatic Speech Recognition
[ "Kartik Audhkhasi", "Tongzhou Chen", "Bhuvana Ramabhadran", "Pedro J. Moreno" ]
https://www.isca-archive.org/interspeech_2021/audhkhasi21_interspeech.html
https://www.isca-archive.org/interspeech_2021/audhkhasi21_interspeech.pdf
10.21437/Interspeech.2021-720
1812-1816
@inproceedings{audhkhasi21_interspeech, title = {{Mixture Model Attention: Flexible Streaming and Non-Streaming Automatic Speech Recognition}}, author = {Kartik Audhkhasi and Tongzhou Chen and Bhuvana Ramabhadran and Pedro J. Moreno}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1...
Streaming automatic speech recognition (ASR) hypothesizes words as soon as the input audio arrives, whereas non-streaming ASR can potentially wait for the completion of the entire utterance to hypothesize words. Streaming and non-streaming ASR systems have typically used different acoustic encoders. Recent work has att...
null
null
inaguma21_interspeech
StableEmit: Selection Probability Discount for Reducing Emission Latency of Streaming Monotonic Attention ASR
[ "Hirofumi Inaguma", "Tatsuya Kawahara" ]
https://www.isca-archive.org/interspeech_2021/inaguma21_interspeech.html
https://www.isca-archive.org/interspeech_2021/inaguma21_interspeech.pdf
10.21437/Interspeech.2021-1110
1817-1821
@inproceedings{inaguma21_interspeech, title = {{StableEmit: Selection Probability Discount for Reducing Emission Latency of Streaming Monotonic Attention ASR}}, author = {Hirofumi Inaguma and Tatsuya Kawahara}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1817--1821}, doi ...
While attention-based encoder-decoder (AED) models have been successfully extended to the online variants for streaming automatic speech recognition (ASR), such as monotonic chunkwise attention (MoChA), the models still have a large label emission latency because of the unconstrained end-to-end training objective. Prev...
2107.00635
title_snapshot
moritz21_interspeech
Dual Causal/Non-Causal Self-Attention for Streaming End-to-End Speech Recognition
[ "Niko Moritz", "Takaaki Hori", "Jonathan Le Roux" ]
https://www.isca-archive.org/interspeech_2021/moritz21_interspeech.html
https://www.isca-archive.org/interspeech_2021/moritz21_interspeech.pdf
10.21437/Interspeech.2021-1693
1822-1826
@inproceedings{moritz21_interspeech, title = {{Dual Causal/Non-Causal Self-Attention for Streaming End-to-End Speech Recognition}}, author = {Niko Moritz and Takaaki Hori and Jonathan Le Roux}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1822--1826}, doi = {10.21437/Inter...
Attention-based end-to-end automatic speech recognition (ASR) systems have recently demonstrated state-of-the-art results for numerous tasks. However, the application of self-attention and attention-based encoder-decoder models remains challenging for streaming ASR, where each word must be recognized shortly after it w...
2107.01269
title_snapshot
kim21d_interspeech
Multi-Mode Transformer Transducer with Stochastic Future Context
[ "Kwangyoun Kim", "Felix Wu", "Prashant Sridhar", "Kyu J. Han", "Shinji Watanabe" ]
https://www.isca-archive.org/interspeech_2021/kim21d_interspeech.html
https://www.isca-archive.org/interspeech_2021/kim21d_interspeech.pdf
10.21437/Interspeech.2021-1953
1827-1831
@inproceedings{kim21d_interspeech, title = {{Multi-Mode Transformer Transducer with Stochastic Future Context}}, author = {Kwangyoun Kim and Felix Wu and Prashant Sridhar and Kyu J. Han and Shinji Watanabe}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1827--1831}, doi = {...
Automatic speech recognition (ASR) models make fewer errors when more surrounding speech information is presented as context. Unfortunately, acquiring a larger future context leads to higher latency. There exists an inevitable trade-off between speed and accuracy. Naïvely, to fit different latency requirements, people ...
2106.09760
title_snapshot
ren21_interspeech
A Causal U-Net Based Neural Beamforming Network for Real-Time Multi-Channel Speech Enhancement
[ "Xinlei Ren", "Xu Zhang", "Lianwu Chen", "Xiguang Zheng", "Chen Zhang", "Liang Guo", "Bing Yu" ]
https://www.isca-archive.org/interspeech_2021/ren21_interspeech.html
https://www.isca-archive.org/interspeech_2021/ren21_interspeech.pdf
10.21437/Interspeech.2021-1457
1832-1836
@inproceedings{ren21_interspeech, title = {{A Causal U-Net Based Neural Beamforming Network for Real-Time Multi-Channel Speech Enhancement}}, author = {Xinlei Ren and Xu Zhang and Lianwu Chen and Xiguang Zheng and Chen Zhang and Liang Guo and Bing Yu}, year = {2021}, booktitle = {{Interspeech 2021}}...
People are meeting through video conferencing more often. While single channel speech enhancement techniques are useful for the individual participants, the speech quality will be significantly degraded in large meeting rooms where the far-field and reverberate conditions are introduced. Approaches based on microphone ...
null
null
zhu21d_interspeech
A Partitioned-Block Frequency-Domain Adaptive Kalman Filter for Stereophonic Acoustic Echo Cancellation
[ "Rui Zhu", "Feiran Yang", "Yuepeng Li", "Shidong Shang" ]
https://www.isca-archive.org/interspeech_2021/zhu21d_interspeech.html
https://www.isca-archive.org/interspeech_2021/zhu21d_interspeech.pdf
10.21437/Interspeech.2021-135
1837-1841
@inproceedings{zhu21d_interspeech, title = {{A Partitioned-Block Frequency-Domain Adaptive Kalman Filter for Stereophonic Acoustic Echo Cancellation}}, author = {Rui Zhu and Feiran Yang and Yuepeng Li and Shidong Shang}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1837--1841}, ...
The rapid development of online video conferencing systems has caused renewed attention to the multi-channel recording and playback systems. Stereophonic acoustic echo cancellation (SAEC) is the key issue of this systems. This paper proposes an optimally designed partitioned-block frequency-domain Kalman filter (PBFDKF...
null
null
wang21p_interspeech
Real-Time Independent Vector Analysis Using Semi-Supervised Nonnegative Matrix Factorization as a Source Model
[ "Taihui Wang", "Feiran Yang", "Rui Zhu", "Jun Yang" ]
https://www.isca-archive.org/interspeech_2021/wang21p_interspeech.html
https://www.isca-archive.org/interspeech_2021/wang21p_interspeech.pdf
10.21437/Interspeech.2021-146
1842-1846
@inproceedings{wang21p_interspeech, title = {{Real-Time Independent Vector Analysis Using Semi-Supervised Nonnegative Matrix Factorization as a Source Model}}, author = {Taihui Wang and Feiran Yang and Rui Zhu and Jun Yang}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1842--1846}...
Online independent vector analysis (IVA) based on auxiliary technology is effective to separate audio source in real time. However, the separated signal may contain residual interference noise because the source model of IVA lacks flexibility and cannot treat the specific harmonic structures of sources. This paper pres...
null
null
han21b_interspeech
Improving Channel Decorrelation for Multi-Channel Target Speech Extraction
[ "Jiangyu Han", "Wei Rao", "Yannan Wang", "Yanhua Long" ]
https://www.isca-archive.org/interspeech_2021/han21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/han21b_interspeech.pdf
10.21437/Interspeech.2021-298
1847-1851
@inproceedings{han21b_interspeech, title = {{Improving Channel Decorrelation for Multi-Channel Target Speech Extraction}}, author = {Jiangyu Han and Wei Rao and Yannan Wang and Yanhua Long}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1847--1851}, doi = {10.21437/Interspe...
Target speech extraction has attracted widespread attention. When microphone arrays are available, the additional spatial information can be helpful in extracting the target speech. We have recently proposed a channel decorrelation (CD) mechanism to extract the inter-channel differential information to enhance the refe...
2106.03113
title_snapshot
liu21f_interspeech
Inplace Gated Convolutional Recurrent Neural Network for Dual-Channel Speech Enhancement
[ "Jinjiang Liu", "Xueliang Zhang" ]
https://www.isca-archive.org/interspeech_2021/liu21f_interspeech.html
https://www.isca-archive.org/interspeech_2021/liu21f_interspeech.pdf
10.21437/Interspeech.2021-899
1852-1856
@inproceedings{liu21f_interspeech, title = {{Inplace Gated Convolutional Recurrent Neural Network for Dual-Channel Speech Enhancement}}, author = {Jinjiang Liu and Xueliang Zhang}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1852--1856}, doi = {10.21437/Interspeech.2021-8...
For dual-channel speech enhancement, it is a promising idea to design an end-to-end model based on the traditional array signal processing guideline and the manifold space of multi-channel signals. We found that the idea above can be effectively implemented by the classical convolutional recurrent neural networks (CRN)...
2107.11968
title_snapshot
raj21_interspeech
SRIB-LEAP Submission to Far-Field Multi-Channel Speech Enhancement Challenge for Video Conferencing
[ "R.G. Prithvi Raj", "Rohit Kumar", "M.K. Jayesh", "Anurenjan Purushothaman", "Sriram Ganapathy", "M.A. Basha Shaik" ]
https://www.isca-archive.org/interspeech_2021/raj21_interspeech.html
https://www.isca-archive.org/interspeech_2021/raj21_interspeech.pdf
10.21437/Interspeech.2021-1111
1857-1861
@inproceedings{raj21_interspeech, title = {{SRIB-LEAP Submission to Far-Field Multi-Channel Speech Enhancement Challenge for Video Conferencing}}, author = {R.G. Prithvi Raj and Rohit Kumar and M.K. Jayesh and Anurenjan Purushothaman and Sriram Ganapathy and M.A. Basha Shaik}, year = {2021}, booktit...
This paper presents the details of the SRIB-LEAP submission to the ConferencingSpeech challenge 2021. The challenge involved the task of multi-channel speech enhancement to improve the quality of far field speech from microphone arrays in a video conferencing room. We propose a two stage method involving a beamformer f...
2106.12763
title_snapshot
xue21_interspeech
Real-Time Multi-Channel Speech Enhancement Based on Neural Network Masking with Attention Model
[ "Cheng Xue", "Weilong Huang", "Weiguang Chen", "Jinwei Feng" ]
https://www.isca-archive.org/interspeech_2021/xue21_interspeech.html
https://www.isca-archive.org/interspeech_2021/xue21_interspeech.pdf
10.21437/Interspeech.2021-2266
1862-1866
@inproceedings{xue21_interspeech, title = {{Real-Time Multi-Channel Speech Enhancement Based on Neural Network Masking with Attention Model}}, author = {Cheng Xue and Weilong Huang and Weiguang Chen and Jinwei Feng}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1862--1866}, doi ...
In this paper, we propose a real-time multi-channel speech enhancement method for noise reduction and dereverberation in far-field environments. The proposed method consists of two components: differential beamforming and mask estimation network. The differential beamforming is employed to suppress the interference sig...
null
null
fohr21_interspeech
BERT-Based Semantic Model for Rescoring N-Best Speech Recognition List
[ "Dominique Fohr", "Irina Illina" ]
https://www.isca-archive.org/interspeech_2021/fohr21_interspeech.html
https://www.isca-archive.org/interspeech_2021/fohr21_interspeech.pdf
10.21437/Interspeech.2021-313
1867-1871
@inproceedings{fohr21_interspeech, title = {{BERT-Based Semantic Model for Rescoring N-Best Speech Recognition List}}, author = {Dominique Fohr and Irina Illina}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1867--1871}, doi = {10.21437/Interspeech.2021-313}, issn =...
This work aims to improve automatic speech recognition (ASR) by modeling long-term semantic relations. We propose to perform this through rescoring the ASR N-best hypotheses list. To achieve this, we propose two deep neural network (DNN) models and combine semantic, acoustic, and linguistic information. Our DNN rescori...
2011.00975
title_judge
benes21_interspeech
Text Augmentation for Language Models in High Error Recognition Scenario
[ "Karel Beneš", "Lukáš Burget" ]
https://www.isca-archive.org/interspeech_2021/benes21_interspeech.html
https://www.isca-archive.org/interspeech_2021/benes21_interspeech.pdf
10.21437/Interspeech.2021-627
1872-1876
@inproceedings{benes21_interspeech, title = {{Text Augmentation for Language Models in High Error Recognition Scenario}}, author = {Karel Beneš and Lukáš Burget}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1872--1876}, doi = {10.21437/Interspeech.2021-627}, issn =...
In this paper, we explore several data augmentation strategies for training of language models for speech recognition. We compare augmentation based on global error statistics with one based on unigram statistics of ASR errors and with label-smoothing and its sampled variant. Additionally, we investigate the stability ...
2011.06056
title_snapshot
gao21b_interspeech
On Sampling-Based Training Criteria for Neural Language Modeling
[ "Yingbo Gao", "David Thulke", "Alexander Gerstenberger", "Khoa Viet Tran", "Ralf Schlüter", "Hermann Ney" ]
https://www.isca-archive.org/interspeech_2021/gao21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/gao21b_interspeech.pdf
10.21437/Interspeech.2021-1067
1877-1881
@inproceedings{gao21b_interspeech, title = {{On Sampling-Based Training Criteria for Neural Language Modeling}}, author = {Yingbo Gao and David Thulke and Alexander Gerstenberger and Khoa Viet Tran and Ralf Schlüter and Hermann Ney}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {18...
As the vocabulary size of modern word-based language models becomes ever larger, many sampling-based training criteria are proposed and investigated. The essence of these sampling methods is that the softmax-related traversal over the entire vocabulary can be simplified, giving speedups compared to the baseline. A prob...
2104.10507
title_snapshot
pylkkonen21_interspeech
Fast Text-Only Domain Adaptation of RNN-Transducer Prediction Network
[ "Janne Pylkkönen", "Antti Ukkonen", "Juho Kilpikoski", "Samu Tamminen", "Hannes Heikinheimo" ]
https://www.isca-archive.org/interspeech_2021/pylkkonen21_interspeech.html
https://www.isca-archive.org/interspeech_2021/pylkkonen21_interspeech.pdf
10.21437/Interspeech.2021-1191
1882-1886
@inproceedings{pylkkonen21_interspeech, title = {{Fast Text-Only Domain Adaptation of RNN-Transducer Prediction Network}}, author = {Janne Pylkkönen and Antti Ukkonen and Juho Kilpikoski and Samu Tamminen and Hannes Heikinheimo}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1882--...
Adaption of end-to-end speech recognition systems to new tasks is known to be challenging. A number of solutions have been proposed which apply external language models with various fusion methods, possibly with a combination of two-pass decoding. Also TTS systems have been used to generate adaptation data for the end-...
2104.11127
title_snapshot
cieri21_interspeech
Using Games to Augment Corpora for Language Recognition and Confusability
[ "Christopher Cieri", "James Fiumara", "Jonathan Wright" ]
https://www.isca-archive.org/interspeech_2021/cieri21_interspeech.html
https://www.isca-archive.org/interspeech_2021/cieri21_interspeech.pdf
10.21437/Interspeech.2021-1611
1887-1891
@inproceedings{cieri21_interspeech, title = {{Using Games to Augment Corpora for Language Recognition and Confusability}}, author = {Christopher Cieri and James Fiumara and Jonathan Wright}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1887--1891}, doi = {10.21437/Interspe...
We present a Game with a Purpose to elicit judgements of the language spoken in short audio clips of broadcast and conversational telephone speech, the resulting corpus and their potential use in research on language recognition and confusability.
null
null
fenu21_interspeech
Fair Voice Biometrics: Impact of Demographic Imbalance on Group Fairness in Speaker Recognition
[ "Gianni Fenu", "Mirko Marras", "Giacomo Medda", "Giacomo Meloni" ]
https://www.isca-archive.org/interspeech_2021/fenu21_interspeech.html
https://www.isca-archive.org/interspeech_2021/fenu21_interspeech.pdf
10.21437/Interspeech.2021-1857
1892-1896
@inproceedings{fenu21_interspeech, title = {{Fair Voice Biometrics: Impact of Demographic Imbalance on Group Fairness in Speaker Recognition}}, author = {Gianni Fenu and Mirko Marras and Giacomo Medda and Giacomo Meloni}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1892--1896}, ...
Speaker recognition systems are playing a key role in modern online applications. Though the susceptibility of these systems to discrimination according to group fairness metrics has been recently studied, their assessment has been mainly focused on the difference in equal error rate across groups, not accounting for o...
null
null
zhang21m_interspeech
Knowledge Distillation from Multi-Modality to Single-Modality for Person Verification
[ "Leying Zhang", "Zhengyang Chen", "Yanmin Qian" ]
https://www.isca-archive.org/interspeech_2021/zhang21m_interspeech.html
https://www.isca-archive.org/interspeech_2021/zhang21m_interspeech.pdf
10.21437/Interspeech.2021-2119
1897-1901
@inproceedings{zhang21m_interspeech, title = {{Knowledge Distillation from Multi-Modality to Single-Modality for Person Verification}}, author = {Leying Zhang and Zhengyang Chen and Yanmin Qian}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1897--1901}, doi = {10.21437/Int...
Voice and face are two important biometric characteristics that can be used for person identity verification. Previous works have proved the strong complementarity between audio and visual modalities in person verification tasks that multi-modality system can achieve significant performance improvement compared to sing...
null
null
noe21_interspeech
Adversarial Disentanglement of Speaker Representation for Attribute-Driven Privacy Preservation
[ "Paul-Gauthier Noé", "Mohammad Mohammadamini", "Driss Matrouf", "Titouan Parcollet", "Andreas Nautsch", "Jean-François Bonastre" ]
https://www.isca-archive.org/interspeech_2021/noe21_interspeech.html
https://www.isca-archive.org/interspeech_2021/noe21_interspeech.pdf
10.21437/Interspeech.2021-1712
1902-1906
@inproceedings{noe21_interspeech, title = {{Adversarial Disentanglement of Speaker Representation for Attribute-Driven Privacy Preservation}}, author = {Paul-Gauthier Noé and Mohammad Mohammadamini and Driss Matrouf and Titouan Parcollet and Andreas Nautsch and Jean-François Bonastre}, year = {2021}, ...
In speech technologies, speaker’s voice representation is used in many applications such as speech recognition, voice conversion, speech synthesis and, obviously, user authentication. Modern vocal representations of the speaker are based on neural embeddings. In addition to the targeted information, these representatio...
2012.04454
title_snapshot
romana21_interspeech
Automatically Detecting Errors and Disfluencies in Read Speech to Predict Cognitive Impairment in People with Parkinson’s Disease
[ "Amrit Romana", "John Bandon", "Matthew Perez", "Stephanie Gutierrez", "Richard Richter", "Angela Roberts", "Emily Mower Provost" ]
https://www.isca-archive.org/interspeech_2021/romana21_interspeech.html
https://www.isca-archive.org/interspeech_2021/romana21_interspeech.pdf
10.21437/Interspeech.2021-1694
1907-1911
@inproceedings{romana21_interspeech, title = {{Automatically Detecting Errors and Disfluencies in Read Speech to Predict Cognitive Impairment in People with Parkinson’s Disease}}, author = {Amrit Romana and John Bandon and Matthew Perez and Stephanie Gutierrez and Richard Richter and Angela Roberts and Emily...
Parkinson’s disease (PD) is a central nervous system disorder that causes motor impairment. Recent studies have found that people with PD also often suffer from cognitive impairment (CI). While a large body of work has shown that speech can be used to predict motor symptom severity in people with PD, much less has focu...
null
null
vaysse21_interspeech
Automatic Extraction of Speech Rhythm Descriptors for Speech Intelligibility Assessment in the Context of Head and Neck Cancers
[ "Robin Vaysse", "Jérôme Farinas", "Corine Astésano", "Régine André-Obrecht" ]
https://www.isca-archive.org/interspeech_2021/vaysse21_interspeech.html
https://www.isca-archive.org/interspeech_2021/vaysse21_interspeech.pdf
10.21437/Interspeech.2021-1736
1912-1916
@inproceedings{vaysse21_interspeech, title = {{Automatic Extraction of Speech Rhythm Descriptors for Speech Intelligibility Assessment in the Context of Head and Neck Cancers}}, author = {Robin Vaysse and Jérôme Farinas and Corine Astésano and Régine André-Obrecht}, year = {2021}, booktitle = {{Inte...
The temporal dimension of speech acoustics is rarely taken into account in automatic models for Speech Intelligibility evaluation, although the rhythmic recurrence of phonemes, syllables and prosodic groups are allegedly good predictors of speech intelligibility. The present study aims at unravelling those automatic pa...
null
null
qi21b_interspeech
Speech Disorder Classification Using Extended Factorized Hierarchical Variational Auto-Encoders
[ "Jinzi Qi", "Hugo Van hamme" ]
https://www.isca-archive.org/interspeech_2021/qi21b_interspeech.html
https://www.isca-archive.org/interspeech_2021/qi21b_interspeech.pdf
10.21437/Interspeech.2021-2180
1917-1921
@inproceedings{qi21b_interspeech, title = {{Speech Disorder Classification Using Extended Factorized Hierarchical Variational Auto-Encoders}}, author = {Jinzi Qi and Hugo {Van hamme}}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1917--1921}, doi = {10.21437/Interspeech.20...
Objective speech disorder classification for speakers with communication difficulty is desirable for diagnosis and administering therapy. With the current state of speech technology, it is evident to propose neural networks for this application. But neural network model training is hampered by a lack of labeled disorde...
2106.07337
title_snapshot
mathad21_interspeech
The Impact of Forced-Alignment Errors on Automatic Pronunciation Evaluation
[ "Vikram C. Mathad", "Tristan J. Mahr", "Nancy Scherer", "Kathy Chapman", "Katherine C. Hustad", "Julie Liss", "Visar Berisha" ]
https://www.isca-archive.org/interspeech_2021/mathad21_interspeech.html
https://www.isca-archive.org/interspeech_2021/mathad21_interspeech.pdf
10.21437/Interspeech.2021-1403
1922-1926
@inproceedings{mathad21_interspeech, title = {{The Impact of Forced-Alignment Errors on Automatic Pronunciation Evaluation}}, author = {Vikram C. Mathad and Tristan J. Mahr and Nancy Scherer and Kathy Chapman and Katherine C. Hustad and Julie Liss and Visar Berisha}, year = {2021}, booktitle = {{Int...
Automatic evaluation of phone-level pronunciation scores typically involves two stages: (1) automatic phonetic segmentation via text-constrained phoneme alignment and (2) quantification of acoustic deviation for each phoneme-level relative to a database of correctly-pronounced speech. It’s clear that the second stage d...
null
null
villatorotello21_interspeech
Late Fusion of the Available Lexicon and Raw Waveform-Based Acoustic Modeling for Depression and Dementia Recognition
[ "Esaú Villatoro-Tello", "S. Pavankumar Dubagunta", "Julian Fritsch", "Gabriela Ramírez-de-la-Rosa", "Petr Motlicek", "Mathew Magimai-Doss" ]
https://www.isca-archive.org/interspeech_2021/villatorotello21_interspeech.html
https://www.isca-archive.org/interspeech_2021/villatorotello21_interspeech.pdf
10.21437/Interspeech.2021-1288
1927-1931
@inproceedings{villatorotello21_interspeech, title = {{Late Fusion of the Available Lexicon and Raw Waveform-Based Acoustic Modeling for Depression and Dementia Recognition}}, author = {Esaú Villatoro-Tello and S. Pavankumar Dubagunta and Julian Fritsch and Gabriela Ramírez-de-la-Rosa and Petr Motlicek and M...
Mental disorders, e.g. depression and dementia, are categorized as priority conditions according to the World Health Organization (WHO). When diagnosing, psychologists employ structured questionnaires/interviews, and different cognitive tests. Although accurate, there is an increasing necessity of developing digital me...
null
null
shandiz21_interspeech
Neural Speaker Embeddings for Ultrasound-Based Silent Speech Interfaces
[ "Amin Honarmandi Shandiz", "László Tóth", "Gábor Gosztolya", "Alexandra Markó", "Tamás Gábor Csapó" ]
https://www.isca-archive.org/interspeech_2021/shandiz21_interspeech.html
https://www.isca-archive.org/interspeech_2021/shandiz21_interspeech.pdf
10.21437/Interspeech.2021-1466
1932-1936
@inproceedings{shandiz21_interspeech, title = {{Neural Speaker Embeddings for Ultrasound-Based Silent Speech Interfaces}}, author = {Amin Honarmandi Shandiz and László Tóth and Gábor Gosztolya and Alexandra Markó and Tamás Gábor Csapó}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = ...
Articulatory-to-acoustic mapping seeks to reconstruct speech from a recording of the articulatory movements, for example, an ultrasound video. Just like speech signals, these recordings represent not only the linguistic content, but are also highly specific to the actual speaker. Hence, due to the lack of multi-speaker...
2106.04552
title_snapshot
lamba21_interspeech
Cross-Modal Learning for Audio-Visual Video Parsing
[ "Jatin Lamba", "Abhishek", "Jayaprakash Akula", "Rishabh Dabral", "Preethi Jyothi", "Ganesh Ramakrishnan" ]
https://www.isca-archive.org/interspeech_2021/lamba21_interspeech.html
https://www.isca-archive.org/interspeech_2021/lamba21_interspeech.pdf
10.21437/Interspeech.2021-2135
1937-1941
@inproceedings{lamba21_interspeech, title = {{Cross-Modal Learning for Audio-Visual Video Parsing}}, author = {Jatin Lamba and Abhishek and Jayaprakash Akula and Rishabh Dabral and Preethi Jyothi and Ganesh Ramakrishnan}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1937--1941}, ...
In this paper, we present a novel approach to the audio-visual video parsing (AVVP) task that demarcates events from a video separately for audio and visual modalities. The proposed parsing approach simultaneously detects the temporal boundaries in terms of start and end times of such events. We show how AVVP can benef...
2104.04598
title_snapshot
cook21_interspeech
A Psychology-Driven Computational Analysis of Political Interviews
[ "Darren Cook", "Miri Zilka", "Simon Maskell", "Laurence Alison" ]
https://www.isca-archive.org/interspeech_2021/cook21_interspeech.html
https://www.isca-archive.org/interspeech_2021/cook21_interspeech.pdf
10.21437/Interspeech.2021-2249
1942-1946
@inproceedings{cook21_interspeech, title = {{A Psychology-Driven Computational Analysis of Political Interviews}}, author = {Darren Cook and Miri Zilka and Simon Maskell and Laurence Alison}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1942--1946}, doi = {10.21437/Intersp...
Can an interviewer influence the cooperativeness of an interviewee? The role of an interviewer in actualising a successful interview is an active field of social psychological research. A large-scale analysis of interviews, however, typically involves time-exorbitant manual tasks and considerable human effort. Despite ...
null
null
santoso21_interspeech
Speech Emotion Recognition Based on Attention Weight Correction Using Word-Level Confidence Measure
[ "Jennifer Santoso", "Takeshi Yamada", "Shoji Makino", "Kenkichi Ishizuka", "Takekatsu Hiramura" ]
https://www.isca-archive.org/interspeech_2021/santoso21_interspeech.html
https://www.isca-archive.org/interspeech_2021/santoso21_interspeech.pdf
10.21437/Interspeech.2021-411
1947-1951
@inproceedings{santoso21_interspeech, title = {{Speech Emotion Recognition Based on Attention Weight Correction Using Word-Level Confidence Measure}}, author = {Jennifer Santoso and Takeshi Yamada and Shoji Makino and Kenkichi Ishizuka and Takekatsu Hiramura}, year = {2021}, booktitle = {{Interspeec...
Emotion recognition is essential for human behavior analysis and possible through various inputs such as speech and images. However, in practical situations, such as in call center analysis, the available information is limited to speech. This leads to the study of speech emotion recognition (SER). Considering the comp...
null
null
silpachai21_interspeech
Effects of Voice Type and Task on L2 Learners’ Awareness of Pronunciation Errors
[ "Alif Silpachai", "Ivana Rehman", "Taylor Anne Barriuso", "John Levis", "Evgeny Chukharev-Hudilainen", "Guanlong Zhao", "Ricardo Gutierrez-Osuna" ]
https://www.isca-archive.org/interspeech_2021/silpachai21_interspeech.html
https://www.isca-archive.org/interspeech_2021/silpachai21_interspeech.pdf
10.21437/Interspeech.2021-701
1952-1956
@inproceedings{silpachai21_interspeech, title = {{Effects of Voice Type and Task on L2 Learners’ Awareness of Pronunciation Errors}}, author = {Alif Silpachai and Ivana Rehman and Taylor Anne Barriuso and John Levis and Evgeny Chukharev-Hudilainen and Guanlong Zhao and Ricardo Gutierrez-Osuna}, year =...
Research suggests learners may improve their second language (L2) pronunciation by imitating voices with similar acoustic profiles. However, previously reported improvements have been in suprasegmentals (prosodic features such as intonation). It remains unclear if voice similarity applies to L2 segmentals (consonants a...
null
null
menshikova21_interspeech
Lexical Entrainment and Intra-Speaker Variability in Cooperative Dialogues
[ "Alla Menshikova", "Daniil Kocharov", "Tatiana Kachkovskaia" ]
https://www.isca-archive.org/interspeech_2021/menshikova21_interspeech.html
https://www.isca-archive.org/interspeech_2021/menshikova21_interspeech.pdf
10.21437/Interspeech.2021-1441
1957-1961
@inproceedings{menshikova21_interspeech, title = {{Lexical Entrainment and Intra-Speaker Variability in Cooperative Dialogues}}, author = {Alla Menshikova and Daniil Kocharov and Tatiana Kachkovskaia}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1957--1961}, doi = {10.214...
In dialogues, intra-speaker variability is often explained by the relationship between interlocutors. A person may speak differently with a friend and a stranger or depending on the interlocutor’s gender or age — in all these cases we expect speech entrainment, but the degree of entrainment may vary. In this research, ...
null
null
nasreen21_interspeech
Detecting Alzheimer’s Disease Using Interactional and Acoustic Features from Spontaneous Speech
[ "Shamila Nasreen", "Julian Hough", "Matthew Purver" ]
https://www.isca-archive.org/interspeech_2021/nasreen21_interspeech.html
https://www.isca-archive.org/interspeech_2021/nasreen21_interspeech.pdf
10.21437/Interspeech.2021-1526
1962-1966
@inproceedings{nasreen21_interspeech, title = {{Detecting Alzheimer’s Disease Using Interactional and Acoustic Features from Spontaneous Speech}}, author = {Shamila Nasreen and Julian Hough and Matthew Purver}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1962--1966}, doi ...
Alzheimer’s Disease (AD) is a form of Dementia that manifests in cognitive decline including memory, language, and changes in behavior. Speech data has proven valuable for inferring cognitive status, used in many health assessment tasks, and can be easily elicited in natural settings. Much work focuses on analysis usin...
null
null
kothare21_interspeech
Investigating the Interplay Between Affective, Phonatory and Motoric Subsystems in Autism Spectrum Disorder Using a Multimodal Dialogue Agent
[ "Hardik Kothare", "Vikram Ramanarayanan", "Oliver Roesler", "Michael Neumann", "Jackson Liscombe", "William Burke", "Andrew Cornish", "Doug Habberstad", "Alaa Sakallah", "Sara Markuson", "Seemran Kansara", "Afik Faerman", "Yasmine Bensidi-Slimane", "Laura Fry", "Saige Portera", "David ...
https://www.isca-archive.org/interspeech_2021/kothare21_interspeech.html
https://www.isca-archive.org/interspeech_2021/kothare21_interspeech.pdf
10.21437/Interspeech.2021-1796
1967-1971
@inproceedings{kothare21_interspeech, title = {{Investigating the Interplay Between Affective, Phonatory and Motoric Subsystems in Autism Spectrum Disorder Using a Multimodal Dialogue Agent}}, author = {Hardik Kothare and Vikram Ramanarayanan and Oliver Roesler and Michael Neumann and Jackson Liscombe and Wi...
We explore the utility of an on-demand multimodal conversational platform in extracting speech and facial metrics in children with Autism Spectrum Disorder (ASD). We investigate the extent to which these metrics correlate with objective clinical measures, particularly as they pertain to the interplay between the affect...
null
null
ishi21_interspeech
Analysis of Eye Gaze Reasons and Gaze Aversions During Three-Party Conversations
[ "Carlos Toshinori Ishi", "Taiken Shintani" ]
https://www.isca-archive.org/interspeech_2021/ishi21_interspeech.html
https://www.isca-archive.org/interspeech_2021/ishi21_interspeech.pdf
10.21437/Interspeech.2021-2134
1972-1976
@inproceedings{ishi21_interspeech, title = {{Analysis of Eye Gaze Reasons and Gaze Aversions During Three-Party Conversations}}, author = {Carlos Toshinori Ishi and Taiken Shintani}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1972--1976}, doi = {10.21437/Interspeech.2021...
The background of this study is the generation of natural gaze behaviors in human-robot multimodal interaction. For that purpose, in this study we analyzed gaze behaviors of multiple speakers in a dataset containing three-party conversations, in terms of the reasons/intentions of their gaze events. Analyses of the gaze...
null
null
kim21e_interspeech
Semantic Distance: A New Metric for ASR Performance Analysis Towards Spoken Language Understanding
[ "Suyoun Kim", "Abhinav Arora", "Duc Le", "Ching-Feng Yeh", "Christian Fuegen", "Ozlem Kalinli", "Michael L. Seltzer" ]
https://www.isca-archive.org/interspeech_2021/kim21e_interspeech.html
https://www.isca-archive.org/interspeech_2021/kim21e_interspeech.pdf
10.21437/Interspeech.2021-1929
1977-1981
@inproceedings{kim21e_interspeech, title = {{Semantic Distance: A New Metric for ASR Performance Analysis Towards Spoken Language Understanding}}, author = {Suyoun Kim and Abhinav Arora and Duc Le and Ching-Feng Yeh and Christian Fuegen and Ozlem Kalinli and Michael L. Seltzer}, year = {2021}, bookt...
Word Error Rate (WER) has been the predominant metric used to evaluate the performance of automatic speech recognition (ASR) systems. However, WER is sometimes not a good indicator for downstream Natural Language Understanding (NLU) tasks, such as intent recognition, slot filling, and semantic parsing in task-oriented ...
2104.02138
title_snapshot
wang21q_interspeech
A Light-Weight Contextual Spelling Correction Model for Customizing Transducer-Based Speech Recognition Systems
[ "Xiaoqiang Wang", "Yanqing Liu", "Sheng Zhao", "Jinyu Li" ]
https://www.isca-archive.org/interspeech_2021/wang21q_interspeech.html
https://www.isca-archive.org/interspeech_2021/wang21q_interspeech.pdf
10.21437/Interspeech.2021-379
1982-1986
@inproceedings{wang21q_interspeech, title = {{A Light-Weight Contextual Spelling Correction Model for Customizing Transducer-Based Speech Recognition Systems}}, author = {Xiaoqiang Wang and Yanqing Liu and Sheng Zhao and Jinyu Li}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1982...
It’s challenging to customize transducer-based automatic speech recognition (ASR) system with context information which is dynamic and unavailable during model training. In this work, we introduce a light-weight contextual spelling correction model to correct context-related recognition errors in transducer-based ASR s...
2108.07493
title_snapshot
shi21_interspeech
Incorporating External POS Tagger for Punctuation Restoration
[ "Ning Shi", "Wei Wang", "Boxin Wang", "Jinfeng Li", "Xiangyu Liu", "Zhouhan Lin" ]
https://www.isca-archive.org/interspeech_2021/shi21_interspeech.html
https://www.isca-archive.org/interspeech_2021/shi21_interspeech.pdf
10.21437/Interspeech.2021-1708
1987-1991
@inproceedings{shi21_interspeech, title = {{Incorporating External POS Tagger for Punctuation Restoration}}, author = {Ning Shi and Wei Wang and Boxin Wang and Jinfeng Li and Xiangyu Liu and Zhouhan Lin}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1987--1991}, doi = {10....
Punctuation restoration is an important post-processing step in automatic speech recognition. Among other kinds of external information, part-of-speech (POS) taggers provide informative tags, suggesting each input token’s syntactic role, which has been shown to be beneficial for the punctuation restoration task. In thi...
2106.06731
title_snapshot
papadourakis21_interspeech
Phonetically Induced Subwords for End-to-End Speech Recognition
[ "Vasileios Papadourakis", "Markus Müller", "Jing Liu", "Athanasios Mouchtaris", "Maurizio Omologo" ]
https://www.isca-archive.org/interspeech_2021/papadourakis21_interspeech.html
https://www.isca-archive.org/interspeech_2021/papadourakis21_interspeech.pdf
10.21437/Interspeech.2021-1787
1992-1996
@inproceedings{papadourakis21_interspeech, title = {{Phonetically Induced Subwords for End-to-End Speech Recognition}}, author = {Vasileios Papadourakis and Markus Müller and Jing Liu and Athanasios Mouchtaris and Maurizio Omologo}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {199...
End-to-end automatic speech recognition systems map a sequence of acoustic features to text. In modern systems, text is encoded to grapheme subwords which are generated by methods designed for text processing tasks and therefore don’t model or take advantage of the statistics of the acoustic features. Here, we present ...
null
null
mansfield21_interspeech
Revisiting Parity of Human vs. Machine Conversational Speech Transcription
[ "Courtney Mansfield", "Sara Ng", "Gina-Anne Levow", "Richard A. Wright", "Mari Ostendorf" ]
https://www.isca-archive.org/interspeech_2021/mansfield21_interspeech.html
https://www.isca-archive.org/interspeech_2021/mansfield21_interspeech.pdf
10.21437/Interspeech.2021-1908
1997-2001
@inproceedings{mansfield21_interspeech, title = {{Revisiting Parity of Human vs. Machine Conversational Speech Transcription}}, author = {Courtney Mansfield and Sara Ng and Gina-Anne Levow and Richard A. Wright and Mari Ostendorf}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {1997...
A number of studies have compared human and machine transcription, showing that automatic speech recognition (ASR) is approaching human performance in some contexts. Most studies look at differences as measured by the standard speech recognition scoring criterion: word error rate (WER). This study looks at more fine-gr...
null
null
huang21f_interspeech
Lookup-Table Recurrent Language Models for Long Tail Speech Recognition
[ "W. Ronny Huang", "Tara N. Sainath", "Cal Peyser", "Shankar Kumar", "David Rybach", "Trevor Strohman" ]
https://www.isca-archive.org/interspeech_2021/huang21f_interspeech.html
https://www.isca-archive.org/interspeech_2021/huang21f_interspeech.pdf
10.21437/Interspeech.2021-340
2002-2006
@inproceedings{huang21f_interspeech, title = {{Lookup-Table Recurrent Language Models for Long Tail Speech Recognition}}, author = {W. Ronny Huang and Tara N. Sainath and Cal Peyser and Shankar Kumar and David Rybach and Trevor Strohman}, year = {2021}, booktitle = {{Interspeech 2021}}, pages ...
We introduce Lookup-Table Language Models (LookupLM), a method for scaling up the size of RNN language models with only a constant increase in the floating point operations, by increasing the expressivity of the embedding table. In particular, we instantiate an (additional) embedding table which embeds the previous n-g...
2104.04552
title_snapshot
andresferrer21_interspeech
Contextual Density Ratio for Language Model Biasing of Sequence to Sequence ASR Systems
[ "Jesús Andrés-Ferrer", "Dario Albesano", "Puming Zhan", "Paul Vozila" ]
https://www.isca-archive.org/interspeech_2021/andresferrer21_interspeech.html
https://www.isca-archive.org/interspeech_2021/andresferrer21_interspeech.pdf
10.21437/Interspeech.2021-443
2007-2011
@inproceedings{andresferrer21_interspeech, title = {{Contextual Density Ratio for Language Model Biasing of Sequence to Sequence ASR Systems}}, author = {Jesús Andrés-Ferrer and Dario Albesano and Puming Zhan and Paul Vozila}, year = {2021}, booktitle = {{Interspeech 2021}}, pages = {2007--201...
End-2-end (E2E) models have become increasingly popular in some ASR tasks because of their performance and advantages. These E2E models directly approximate the posterior distribution of tokens given the acoustic inputs. Consequently, the E2E systems implicitly define a language model (LM) over the output tokens, which...
2206.14623
title_snapshot