arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2002.06312eess.AScs.LGcs.SD
收藏

端到端语音识别神经网络训练的小能量掩蔽

Small energy masking for improved neural network training for end-to-end speech recognition

Samsung Research(三星研究院)

Chanwoo Kim, Kwangyoun Kim, Sathish Reddy Indurthi

摘要:

在本文中,我们提出了一个小能量屏蔽(SEM)算法,它屏蔽输入值低于一定的阈值。更具体地,如果时频仓中的滤波器组能量小于某个能量阈值,则该时频仓被掩蔽。采用均匀分布来随机地生成该能量阈值与以分贝为单位的每个发声的峰值滤波器组能量的比率。未掩蔽的特征元素被缩放,使得特征值的总和通过该掩蔽过程保持相同。这个非常简单的算法显示出相对于基线端到端语音识别系统,标准LibriSpeech测试-干净和测试-其他集的字错误率(WER)分别提高了11.2%和13.5%。此外,与输入丢弃算法相比,SEM算法在相同的LibriSpeech测试干净集和测试其他集上显示出相对7.7%和11.6%的改进。通过使用Transformer LM的改进的浅融合技术,我们在LibriSpeech测试-干净集上获得了2.62%的WER,在LibriSpeech测试-其他集上获得了7.87%的WER。

英文摘要:

In this paper, we present a Small Energy Masking (SEM) algorithm, which masks inputs having values below a certain threshold. More specifically, a time-frequency bin is masked if the filterbank energy in this bin is less than a certain energy threshold. A uniform distribution is employed to randomly generate the ratio of this energy threshold to the peak filterbank energy of each utterance in decibels. The unmasked feature elements are scaled so that the total sum of the feature values remain the same through this masking procedure. This very simple algorithm shows relatively 11.2 % and 13.5 % Word Error Rate (WER) improvements on the standard LibriSpeech test-clean and test-other sets over the baseline end-to-end speech recognition system. Additionally, compared to the input dropout algorithm, SEM algorithm shows relatively 7.7 % and 11.6 % improvements on the same LibriSpeech test-clean and test-other sets. With a modified shallow-fusion technique with a Transformer LM, we obtained a 2.62 % WER on the LibriSpeech test-clean set and a 7.87 % WER on the LibriSpeech test-other set.

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。

↑