摘要
本文介绍了一种利用复倒谱来实现气声发音重建的方法。首先分析了气声发音的语音特征；进而在复倒谱序列中加入基频特征使其恢复到正常的语音。对元音［a］以及实际语音段进行了处理，均有较好的效果。
This paper intreduces a method which uses complex cepstrum to restore un- voiced speech. First, the characteristics of the unvoiced speech is analyzed; then, the fundamental frequency is added to the complex cepstrum to restore it to the normal speech. We have processed vowel [a] and actual speech segments. The restorations worked effecitvely.
出处
《应用声学》
CSCD
1996年第5期41-44,共4页
Journal of Applied Acoustics
关键词
复倒谱
气声发音
基频
录音
Complex cepstrum,Unvoiced speech,Fundamental frequency
