Silence Segment Detection of English Speech Signals
Zhaoyin CHU, Shaoqin Pan, Jinglan Liu, Liangliang Zhao · 2015
ABSTRCT: To differentiate between silence and voice segment of English speech signal is very important in English IBT systems.In order to improve the accuracy of the end detection in traditional detection method of speech end, a silence/voice detection method of speech signals using wavelet transform parameters is proposed in this paper.Based on high performance of frequency segmentation and energy focusing of the wavelet, the statistical parameters of the speech signals on different sub bands are extracted and the effectiveness of this method is validated by simulation experiment results under different noise conditions.