Publication Details

Eat: Enhanced ASR-TTS for Self-Supervised Speech Recognition

BASKAR, M.; BURGET, L.; WATANABE, S.; ASTUDILLO, R.; ČERNOCKÝ, J. Eat: Enhanced ASR-TTS for Self-Supervised Speech Recognition. In ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). Toronto, Ontario: IEEE Signal Processing Society, 2021. p. 6753-6757. ISBN: 978-1-7281-7605-5.
Czech title
EAT: Obohacený systém ASR-TTS pro samoučící se rozpoznávání řeči
Type
conference paper
Language
English
Authors
Baskar Murali Karthick, Ing., Ph.D.
Burget Lukáš, doc. Ing., Ph.D. (DCGM)
Watanabe Shinji (FIT)
ASTUDILLO, R.
Černocký Jan, prof. Dr. Ing. (DCGM)
URL
Keywords

cycle-consistency, self-supervision, sequence-tosequence, speech recognition

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.

Published
2021
Pages
6753–6757
Proceedings
ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
ISBN
978-1-7281-7605-5
Publisher
IEEE Signal Processing Society
Place
Toronto, Ontario
DOI
UT WoS
000704288407006
EID Scopus
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"
}
Back to top