Pretraining by backtranslation for end-to-end ASR in low-resource settings

Matthew Wiesner, Adithya Renduchintala, Shinji Watanabe, Chunxi Liu, Najim Dehak, Sanjeev Khudanpur

研究成果: Conference article査読

2 被引用数 (Scopus)

抄録

We explore training attention-based encoder-decoder ASR in low-resource settings. These models perform poorly when trained on small amounts of transcribed speech, in part because they depend on having sufficient target-side text to train the attention and decoder networks. In this paper we address this shortcoming by pretraining our network parameters using only text-based data and transcribed speech from other languages. We analyze the relative contributions of both sources of data. Across 3 test languages, our text-based approach resulted in a 20% average relative improvement over a text-based augmentation technique without pretraining. Using transcribed speech from nearby languages gives a further 20-30% relative reduction in character error rate.

本文言語English
ページ(範囲)4375-4379
ページ数5
ジャーナルProceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH
2019-September
DOI
出版ステータスPublished - 2019
外部発表はい
イベント20th Annual Conference of the International Speech Communication Association: Crossroads of Speech and Language, INTERSPEECH 2019 - Graz, Austria
継続期間: 2019 9 152019 9 19

ASJC Scopus subject areas

  • 言語および言語学
  • 人間とコンピュータの相互作用
  • 信号処理
  • ソフトウェア
  • モデリングとシミュレーション

フィンガープリント

「Pretraining by backtranslation for end-to-end ASR in low-resource settings」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル