arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2102.01930cs.SDcs.LGeess.AS
收藏

基于自监督多粒度框架的通用语音表示学习

General-Purpose Speech Representation Learning through a Self-Supervised Multi-Granularity Framework

University of Science and Technology of China(中国科学技术大学);Microsoft Research Asia(微软亚洲研究院)

Yucheng Zhao, Dacheng Yin, Chong Luo, Zhiyuan Zhao, Chuanxin Tang, Wenjun Zeng, Zheng-Jun Zha

摘要:

本文提出了一个自监督学习框架,命名为MGF,通用语音表示学习。在MGF的设计中,考虑了语音的层次性。具体来说,我们建议使用生成学习方法在小时间尺度上捕获细粒度信息,并使用判别学习方法在大时间尺度上提取粗粒度或语义信息。对于音素级学习,我们借用了掩蔽语言模型的思想,但通过用对比损失代替分类损失来为连续语音信号定制它。我们证实了我们的设计,通过评估MGF表示在各种下游任务,包括音素分类,说话人分类,语音识别,和情感分类。实验表明,在不同的时间尺度上进行训练需要不同的训练目标和损失函数,这在总体上是相辅相成的,从而导致更好的性能。

英文摘要:

This paper presents a self-supervised learning framework, named MGF, for general-purpose speech representation learning. In the design of MGF, speech hierarchy is taken into consideration. Specifically, we propose to use generative learning approaches to capture fine-grained information at small time scales and use discriminative learning approaches to distill coarse-grained or semantic information at large time scales. For phoneme-scale learning, we borrow idea from the masked language model but tailor it for the continuous speech signal by replacing classification loss with a contrastive loss. We corroborate our design by evaluating MGF representation on various downstream tasks, including phoneme classification, speaker classification, speech recognition, and emotion classification. Experiments verify that training at different time scales needs different training targets and loss functions, which in general complement each other and lead to a better performance.

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。

↑