Forward Translation to Mix Data for Speech Translation

Zhipeng Wang, Hongjing Xu, Shuoying Chen, Yuhang Guo*

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

End-to-End speech translation means that using a model to translate speech in one language into text in another language. Currently, the main challenge in the field of speech translation is data scarcity. Existing works solve this problem by using text information or applying data augmentation. However, these works only focus on the exploitation of a single corpus, ignoring the full use of existing human-labeled different-sources data. In this paper, we introduce a simple method to solve the data scarcity problem: training a model with simply mixed data and applying the forward translation method to expand the training set. We perform experiments on covost v2 French-English and mTEDx French-English. Our experiments demonstrate that combining the mixture of speech translation corpora with forward translation can yield a better result than the method without mixing.

Original languageEnglish
Title of host publicationICIAI 2023 - 7th International Conference on Innovation in Artificial Intelligence
PublisherAssociation for Computing Machinery
Pages178-182
Number of pages5
ISBN (Electronic)9781450398398
DOIs
Publication statusPublished - 3 Mar 2023
Event7th International Conference on Innovation in Artificial Intelligence, ICIAI 2023 - Harbin, China
Duration: 3 Mar 20235 Mar 2023

Publication series

NameACM International Conference Proceeding Series

Conference

Conference7th International Conference on Innovation in Artificial Intelligence, ICIAI 2023
Country/TerritoryChina
CityHarbin
Period3/03/235/03/23

Keywords

  • Data scarcity
  • Domain adaption
  • Forward-translation
  • Speech translation

Fingerprint

Dive into the research topics of 'Forward Translation to Mix Data for Speech Translation'. Together they form a unique fingerprint.

Cite this