Přístupnostní navigace
E-application
Search Search Close
Publication detail
BASKAR, M. BURGET, L. WATANABE, S. ASTUDILLO, R. ČERNOCKÝ, J.
Original Title
Eat: Enhanced ASR-TTS for Self-Supervised Speech Recognition
Type
conference paper
Language
English
Original Abstract
Self-supervised ASR-TTS models suffer in out-of-domain data conditions. Here we propose an enhanced ASR-TTS (EAT) model that incorporates two main features: 1) The ASR!TTS direction is equipped with a language model reward to penalize the ASR hypotheses before forwarding it to TTS. 2) In the TTS!ASR direction, a hyper-parameter is introduced to scale the attention context from synthesized speech before sending it to ASR to handle out-ofdomain data. Training strategies and the effectiveness of the EAT model are explored under out-of-domain data conditions. The results show that EAT reduces the performance gap between supervised and self-supervised training significantly by absolute 2.6% and 2.7% on Librispeech and BABEL respectively.
Keywords
cycle-consistency, self-supervision, sequence-tosequence, speech recognition
Authors
BASKAR, M.; BURGET, L.; WATANABE, S.; ASTUDILLO, R.; ČERNOCKÝ, J.
Released
6. 6. 2021
Publisher
IEEE Signal Processing Society
Location
Toronto, Ontario
ISBN
978-1-7281-7605-5
Book
ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
Pages from
6753
Pages to
6757
Pages count
5
URL
https://ieeexplore.ieee.org/document/9413375
BibTex
@inproceedings{BUT175793, author="BASKAR, M. and BURGET, L. and WATANABE, S. and ASTUDILLO, R. and ČERNOCKÝ, J.", title="Eat: Enhanced ASR-TTS for Self-Supervised Speech Recognition", booktitle="ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)", year="2021", pages="6753--6757", publisher="IEEE Signal Processing Society", address="Toronto, Ontario", doi="10.1109/ICASSP39728.2021.9413375", isbn="978-1-7281-7605-5", url="https://ieeexplore.ieee.org/document/9413375" }