EXPLORING TRADEOFFS IN MODELS FOR LOW-LATENCY SPEECH ENHANCEMENT

被引：0

作者：

Wilson, Kevin ^{[1
]}

Chinen, Michael ^{[1
]}

Thorpe, Jeremy ^{[1
]}

Patton, Brian ^{[1
]}

Hershey, John ^{[1
]}

Saurous, Rif A. ^{[1
]}

Skoglund, Jan ^{[1
]}

Lyon, Richard F. ^{[1
]}

机构：

[1] Google Res, Google Chrome Audio, Mountain View, CA 94043 USA

来源：

2018 16TH INTERNATIONAL WORKSHOP ON ACOUSTIC SIGNAL ENHANCEMENT (IWAENC) | 2018年

关键词：

speech enhancement; low-latency inference;

D O I：

暂无

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

We explore a variety of neural networks configurations for one- and two-channel spectrogram-mask-based speech enhancement. Our best model improves on previous state-of-the-art performance on the CHiME2 speech enhancement task by 0.4 decibels in signal-to-distortion ratio (SDR). We examine trade-offs such as non-causal look-ahead, computation, and parameter count versus enhancement performance and find that zero-look-ahead models can achieve, on average, within 0.03 dB SDR of our best bidirectional model. Further, we find that 200 milliseconds of look-ahead is sufficient to achieve equivalent performance to our best bidirectional model.

引用

页码：366 / 370

页数：5

共 50 条

[1] A Survey on Low-Latency DNN-Based Speech Enhancement
Drgas, Szymon
[J]. SENSORS, 2023, 23 (03)
[2] Efficient Low-Latency Speech Enhancement with Mobile Audio Streaming Networks
Romaniuk, Michal
Masztalski, Piotr
Piaskowski, Karol
Matuszewski, Mateusz
[J]. INTERSPEECH 2020, 2020, : 3296 - 3300
[3] Low-Latency Neural Speech Translation
Niehues, Jan
Ngoc-Quan Pham
Thanh-Le Ha
Sperber, Matthias
Waibel, Alex
[J]. 19TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION (INTERSPEECH 2018), VOLS 1-6: SPEECH RESEARCH FOR EMERGING MARKETS IN MULTILINGUAL SOCIETIES, 2018, : 1293 - 1297
[4] Low-latency monaural speech enhancement with deep filter-bank equalizer
Zheng, Chengshi
Liu, Wenzhe
Li, Andong
Ke, Yuxuan
Li, Xiaodong
[J]. JOURNAL OF THE ACOUSTICAL SOCIETY OF AMERICA, 2022, 151 (05): : 3291 - 3304
[5] LOW-LATENCY DEEP CLUSTERING FOR SPEECH SEPARATION
Wang, Shanshan
Naithani, Gaurav
Virtanen, Tuomas
[J]. 2019 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP), 2019, : 76 - 80
[6] Dynamic Transcription for Low-latency Speech Translation
Niehues, Jan
Nguyen, Thai Son
Cho, Eunah
Ha, Thanh-Le
Kilgour, Kevin
Mueller, Markus
Sperber, Matthias
Stueker, Sebastian
Waibel, Alex
[J]. 17TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION (INTERSPEECH 2016), VOLS 1-5: UNDERSTANDING SPEECH PROCESSING IN HUMANS AND MACHINES, 2016, : 2513 - 2517
[7] Implementation of low-latency electrolaryngeal speech enhancement based on multi-task CLDNN
Kobayashi, Kazuhiro
Toda, Tomoki
[J]. 28TH EUROPEAN SIGNAL PROCESSING CONFERENCE (EUSIPCO 2020), 2021, : 396 - 400
[8] Amortized Neural Networks for Low-Latency Speech Recognition
Macoskey, Jonathan
Strimel, Grant P.
Su, Jinru
Rastrow, Ariya
[J]. INTERSPEECH 2021, 2021, : 4558 - 4562
[9] Design of a robust MVDR beamforming method with Low-Latency by reconstructing covariance matrix for speech enhancement
Zhou, Jing
Bao, Changchun
Zhang, Xu
Xiong, Wenmeng
Jia, Maoshen
[J]. APPLIED ACOUSTICS, 2023, 211
[10] Improving Low-Latency Mono-Channel Speech Enhancement by Compensation Windows in STFT Analysis
Bui, Minh N.
Tran, Dung N.
Koishida, Kazuhito
Tran, Trac D.
Chin, Peter
[J]. COMPLEX NETWORKS & THEIR APPLICATIONS XII, VOL 1, COMPLEX NETWORKS 2023, 2024, 1141 : 363 - 373

← 1 2 3 4 5 →