摘要
环境鲁棒性是语音识别系统的一个很重要的问题。本文讨论了环境加性噪声和卷积噪声对语音数据的影响 ,以及提高系统的鲁棒性的盲的环境补偿方法。。
Environmental robustness is a very important issue in the field of automatic speech recognition(ASR)research.This paper describes the effect of additive noise and convolution noise on speech data and gives a blind environmental compensation method to improve environmental robustness..
出处
《计算机工程与科学》
CSCD
2000年第4期100-102,共3页
Computer Engineering & Science
关键词
语音识别系统
环境鲁棒性
盲的环境补偿法
environmental robustness
additive noise
convolution noise
blind environmental compensation method
