Publication Details
Semi-supervised Sequence-to-sequence ASR using Unpaired Speech and Text
Watanabe Shinji
ASTUDILLO, R.
HORI, T.
Burget Lukáš, doc. Ing., Ph.D. (DCGM)
Černocký Jan, prof. Dr. Ing. (DCGM)
Sequence-to-sequence, end-to-end, ASR, TTS, semi-supervised, unsupervised, cycle
consistency
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.
@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"
}