Publication detail

Semi-supervised Sequence-to-sequence ASR using Unpaired Speech and Text

BASKAR, M. WATANABE, S. ASTUDILLO, R. HORI, T. BURGET, L. ČERNOCKÝ, J.

Original Title

Semi-supervised Sequence-to-sequence ASR using Unpaired Speech and Text

Type

conference paper

Language

English

Original Abstract

Sequence-to-sequence automatic speech recognition (ASR) models require large quantities of data to attain high performance. For this reason, there has been a recent surge in interest for unsupervised and semi-supervised training in such models. This work builds upon recent results showing notable improvements in semi-supervised training using cycle-consistency and related techniques. Such techniques derive training procedures and losses able to leverage unpaired speech and/or text data by combining ASR with Text-to-Speech (TTS) models. In particular, this work proposes a new semi-supervised loss combining an end-to-end differentiable ASR!TTS loss with TTS!ASR loss. The method is able to leverage both unpaired speech and text data to outperform recently proposed related techniques in terms of %WER. We provide extensive results analyzing the impact of data quantity and speech and text modalities and show consistent gains across WSJ and Librispeech corpora. Our code is provided in ESPnet to reproduce the experiments.

Keywords

Sequence-to-sequence, end-to-end, ASR, TTS, semi-supervised, unsupervised, cycle consistency

Authors

BASKAR, M.; WATANABE, S.; ASTUDILLO, R.; HORI, T.; BURGET, L.; ČERNOCKÝ, J.

Released

15. 9. 2019

Publisher

International Speech Communication Association

Location

Graz

ISBN

1990-9772

Periodical

Proceedings of Interspeech

Year of study

2019

Number

9

State

French Republic

Pages from

3790

Pages to

3794

Pages count

5

URL

https://www.isca-speech.org/archive/Interspeech_2019/pdfs/3167.pdf

BibTex

@inproceedings{BUT159996,
  author="BASKAR, M. and WATANABE, S. and ASTUDILLO, R. and HORI, T. and BURGET, L. and ČERNOCKÝ, J.",
  title="Semi-supervised Sequence-to-sequence ASR using Unpaired Speech and Text",
  booktitle="Proceedings of Interspeech",
  year="2019",
  journal="Proceedings of Interspeech",
  volume="2019",
  number="9",
  pages="3790--3794",
  publisher="International Speech Communication Association",
  address="Graz",
  doi="10.21437/Interspeech.2019-3167",
  issn="1990-9772",
  url="https://www.isca-speech.org/archive/Interspeech_2019/pdfs/3167.pdf"
}

Documents

baskar_is2019_193167.pdf