arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2005.09629eess.AScs.LG
收藏

改进的自动语音识别噪声学生训练算法

Improved Noisy Student Training for Automatic Speech Recognition

Google Inc.(谷歌公司)

Daniel S. Park, Yu Zhang, Ye Jia, Wei Han, Chung-Cheng Chiu, Bo Li, Yonghui Wu, Quoc V. Le

摘要:

最近,一种被称为“噪声学生训练”的半监督学习方法已被证明可以显着提高深度网络的图像分类性能。噪声学生训练是一种迭代的自我训练方法,它利用增强来提高网络性能。在这项工作中,我们适应和改进嘈杂的学生训练自动语音识别,采用(自适应)SpecAugment作为增强方法。我们找到了有效的方法来过滤,平衡和增强自训练迭代之间生成的数据。通过这样做,我们能够在干净/嘈杂的LibriSpeech测试集上获得4.2%/8.6%的单词错误率(WER),只使用LibriSpeech的干净100 h子集作为监督集,其余(860 h)作为未标记集。此外,通过使用LibriLight的unlab-60 k子集作为LibriSpeech 960 h的未标记集,我们能够在干净/嘈杂的LibriSpeech测试集上实现WER 1.7%/3.4%。因此,我们能够改进之前在LibriSpeech 100 h(4.74%/12.20%)和LibriSpeech(1.9%/4.1%)上实现的最先进的清洁/噪声测试WER。

英文摘要:

Recently, a semi-supervised learning method known as "noisy student training" has been shown to improve image classification performance of deep networks significantly. Noisy student training is an iterative self-training method that leverages augmentation to improve network performance. In this work, we adapt and improve noisy student training for automatic speech recognition, employing (adaptive) SpecAugment as the augmentation method. We find effective methods to filter, balance and augment the data generated in between self-training iterations. By doing so, we are able to obtain word error rates (WERs) 4.2%/8.6% on the clean/noisy LibriSpeech test sets by only using the clean 100h subset of LibriSpeech as the supervised set and the rest (860h) as the unlabeled set. Furthermore, we are able to achieve WERs 1.7%/3.4% on the clean/noisy LibriSpeech test sets by using the unlab-60k subset of LibriLight as the unlabeled set for LibriSpeech 960h. We are thus able to improve upon the previous state-of-the-art clean/noisy test WERs achieved on LibriSpeech 100h (4.74%/12.20%) and LibriSpeech (1.9%/4.1%).

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。

↑