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 2310.17558cs.CLcs.LGcs.SDeess.AS
收藏

迈向匹配音素与语音表征

Towards Matching Phones and Speech Representations

University of Edinburgh(爱丁堡大学)

Gene-Ping Yang, Hao Tang

AI总结:

本文将音素类型学习形式化为自监督表征聚类中心与音素嵌入的匹配问题,利用匹配结果生成伪标签并提出新损失函数,与 APC、CPC 联合训练后显著提升下游音素分类性能。

摘要:

从音素实例中学习音素类型一直是一个长期存在的问题,至今仍未解决。在本工作中,我们在自监督学习的背景下重新审视这一问题,并将其形式化为将聚类中心与音素嵌入进行匹配的问题。我们研究了两项使匹配成为可能的关键性质,即自监督表征的聚类中心是否降低了音素实例的变异性,以及是否尊重音素之间的关系。随后,我们利用匹配结果生成伪标签,并引入一种新的损失函数以改进自监督表征。我们的实验表明,匹配结果能够捕捉音素之间的关系。将新的损失函数与常规自监督损失(如 APC 和 CPC)联合训练,可显著提升下游音素分类性能。

英文摘要:

Learning phone types from phone instances has been a long-standing problem, while still being open. In this work, we revisit this problem in the context of self-supervised learning, and pose it as the problem of matching cluster centroids to phone embeddings. We study two key properties that enable matching, namely, whether cluster centroids of self-supervised representations reduce the variability of phone instances and respect the relationship among phones. We then use the matching result to produce pseudo-labels and introduce a new loss function for improving self-supervised representations. Our experiments show that the matching result captures the relationship among phones. Training the new loss function jointly with the regular self-supervised losses, such as APC and CPC, significantly improves the downstream phone classification.

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。

↑