arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2402.01945cs.CL
收藏

端到端语音翻译过滤案例研究

A Case Study on Filtering for End-to-End Speech Translation

George Mason University(乔治梅森大学);Archimedes AI Research Unit, RC Athena(阿基米德人工智能研究单元,RC雅典娜(希腊))

Md Mahfuz Ibn Alam, Antonios Anastasopoulos

AI总结:

本研究通过案例证明,简单过滤技术可有效缩减嘈杂的大规模平行语料,提升端到端语音翻译模型性能,平均BLEU提高4.65。

摘要:

对于任何机器学习任务(如语音到文本或语音到语音翻译),挖掘大规模平行语料库相对容易。尽管这些挖掘出的语料库规模庞大,但其质量却存疑。本研究表明,最简单的过滤技术就能将这些庞大且嘈杂的数据集缩减为更易管理、更干净的数据集。我们还表明,使用这一干净数据集可以提升模型性能,例如在多语言到英语的语音翻译(ST)模型中,平均获得了4.65个BLEU分数的提升。

英文摘要:

It is relatively easy to mine a large parallel corpus for any machine learning task, such as speech-to-text or speech-to-speech translation. Although these mined corpora are large in volume, their quality is questionable. This work shows that the simplest filtering technique can trim down these big, noisy datasets to a more manageable, clean dataset. We also show that using this clean dataset can improve the model's performance, as in the case of the multilingual-to-English Speech Translation (ST) model, where, on average, we obtain a 4.65 BLEU score improvement.

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。

↑