BIT’s System for Multilingual Track

Zhipeng Wang, Yuhang Guo*, Shuoying Chen

*此作品的通讯作者

科研成果: 书/报告/会议事项章节会议稿件同行评审

1 引用 (Scopus)

摘要

This paper describes the system we submitted to the IWSLT 2023 multilingual speech translation track, with the input is speech from one language, and the output is text from 10 target languages. Our system consists of CNN and Transformer, convolutional neural networks downsample speech features and extract local information, while transformer extract global features and output the final results. In our system, we use speech recognition tasks to pre-train encoder parameters, and then use speech translation corpus to train the multilingual speech translation model. We have also adopted other methods to optimize the model, such as data augmentation, model ensemble, etc. Our system can obtain satisfactory results on test sets of 10 languages in the MUST-C corpus.

源语言英语
主期刊名20th International Conference on Spoken Language Translation, IWSLT 2023 - Proceedings of the Conference
编辑Elizabeth Salesky, Marcello Federico, Marine Carpuat
出版商Association for Computational Linguistics
455-460
页数6
ISBN(电子版)9781959429845
出版状态已出版 - 2023
活动20th International Conference on Spoken Language Translation, IWSLT 2023 - Hybrid, Toronto, 加拿大
期限: 13 7月 202314 7月 2023

出版系列

姓名20th International Conference on Spoken Language Translation, IWSLT 2023 - Proceedings of the Conference

会议

会议20th International Conference on Spoken Language Translation, IWSLT 2023
国家/地区加拿大
Hybrid, Toronto
时期13/07/2314/07/23

指纹

探究 'BIT’s System for Multilingual Track' 的科研主题。它们共同构成独一无二的指纹。

引用此