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 2210.15445eess.AScs.CLcs.LGcs.SD
收藏

高效利用大型预训练模型用于低资源自动语音识别(ASR)

Efficient Utilization of Large Pre-Trained Models for Low Resource ASR

RWTH Aachen University(亚琛工业大学);AppTek GmbH(AppTek有限公司)

Peter Vieting, Christoph Lüscher, Julian Dierkes, Ralf Schlüter, Hermann Ney

AI总结:

针对越南语和德语医疗电话低资源ASR,本文研究大型预训练模型的高效利用,结合无监督技术、带宽迁移、架构与训练改进及少量领域内适配数据,显著超越项目基线。

摘要:

无监督表征学习最近帮助自动语音识别(ASR)应对标注数据有限的任务。在此背景下,硬件限制和应用场景引出了一个问题:如何高效利用大型预训练模型并降低其复杂度。在本工作中,我们研究了一个具有挑战性的低资源会电话音语音语料库,其来自越南语和德语的医疗领域。我们展示了在对大型预训练模型进行简单微调之外使用无监督技术的收益,讨论了如何使这些技术适应实际电话任务(包括带宽迁移),并研究了预训练和微调的不同数据条件。通过使用预训练技术,我们相对项目基线取得了22%的提升。通过改进架构和训练,可进一步获得29%的收益;通过加入0.8小时的领域内适配数据,还可获得6%的提升。

英文摘要:

Unsupervised representation learning has recently helped automatic speech recognition (ASR) to tackle tasks with limited labeled data. Following this, hardware limitations and applications give rise to the question how to take advantage of large pre-trained models efficiently and reduce their complexity. In this work, we study a challenging low resource conversational telephony speech corpus from the medical domain in Vietnamese and German. We show the benefits of using unsupervised techniques beyond simple fine-tuning of large pre-trained models, discuss how to adapt them to a practical telephony task including bandwidth transfer and investigate different data conditions for pre-training and fine-tuning. We outperform the project baselines by 22% relative using pretraining techniques. Further gains of 29% can be achieved by refinements of architecture and training and 6% by adding 0.8 h of in-domain adaptation data.

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。

↑