arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2107.02444cs.CL
收藏

IWSLT 2021离线任务的NiuTrans端到端语音翻译系统

The NiuTrans End-to-End Speech Translation System for IWSLT 2021 Offline Task

Northeastern University(东北大学);NiuTrans Research(小牛翻译研究)

Chen Xu, Xiaoqian Liu, Xiaowen Liu, Laohu Wang, Canan Huang, Tong Xiao, Jingbo Zhu

摘要:

本文介绍了NiuTrans端到端语音翻译系统为IWITH 2021离线任务提交的内容,该系统直接将英语音频翻译为德语文本,无需中间转录。我们使用基于Transformer的模型架构,并通过Conformer,相对位置编码以及堆叠的声学和文本编码对其进行增强。为了增加训练数据,英语译文被翻译成德语译文。最后,我们采用集成解码来整合来自不同数据集训练的几个模型的预测。结合这些技术,我们在MuST-C En-De测试集上获得了33.84个BLEU点,这显示了端到端模型的巨大潜力。

英文摘要:

This paper describes the submission of the NiuTrans end-to-end speech translation system for the IWSLT 2021 offline task, which translates from the English audio to German text directly without intermediate transcription. We use the Transformer-based model architecture and enhance it by Conformer, relative position encoding, and stacked acoustic and textual encoding. To augment the training data, the English transcriptions are translated to German translations. Finally, we employ ensemble decoding to integrate the predictions from several models trained with the different datasets. Combining these techniques, we achieve 33.84 BLEU points on the MuST-C En-De test set, which shows the enormous potential of the end-to-end model.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑