arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2005.06065eess.AScs.CLcs.SD
收藏

语音中辅音可懂度的自动估计

Automatic Estimation of Intelligibility Measure for Consonants in Speech

University of Illinois at Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校)

Ali Abavisani, Mark Hasegawa-Johnson

摘要:

在这篇文章中,我们提供了一个模型来估计一个真正的价值衡量个别语音段的可懂度。我们基于卷积神经网络(CNN)训练回归模型,用于与元音\textipa{/A/}相关的塞音\textipa{/p,t,k,b,d,g/},以估计相应的信噪比(SNR),在该信噪比下,辅音-元音(CV)声音对于正常听力(NH)耳朵变得可理解。每个声音的可懂度度量称为SNR $_{90}$,定义为人类参与者能够平均至少90%正确识别辅音的SNR水平,如之前对NH受试者的实验所确定的。将CNN的性能与基于自动语音识别(ASR)的基线预测进行比较,具体地,从SNR中减去恒定偏移,在该恒定偏移处,ASR变得能够正确地标记辅音。与基线相比,我们的模型能够准确估计SNR $_{90}$ ~可懂度测量,平均均方误差(MSE)小于2 [dB $^2$ ],而基线ASR定义的测量计算SNR $_{90}$ ~,方差为5.2至26.6 [dB $^2$ ],具体取决于辅音。

英文摘要:

In this article, we provide a model to estimate a real-valued measure of the intelligibility of individual speech segments. We trained regression models based on Convolutional Neural Networks (CNN) for stop consonants \textipa{/p,t,k,b,d,g/} associated with vowel \textipa{/A/}, to estimate the corresponding Signal to Noise Ratio (SNR) at which the Consonant-Vowel (CV) sound becomes intelligible for Normal Hearing (NH) ears. The intelligibility measure for each sound is called SNR$_{90}$, and is defined to be the SNR level at which human participants are able to recognize the consonant at least 90\% correctly, on average, as determined in prior experiments with NH subjects. Performance of the CNN is compared to a baseline prediction based on automatic speech recognition (ASR), specifically, a constant offset subtracted from the SNR at which the ASR becomes capable of correctly labeling the consonant. Compared to baseline, our models were able to accurately estimate the SNR$_{90}$~intelligibility measure with less than 2 [dB$^2$] Mean Squared Error (MSE) on average, while the baseline ASR-defined measure computes SNR$_{90}$~with a variance of 5.2 to 26.6 [dB$^2$], depending on the consonant.

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。

↑