A Speech Endpoint Detection Method Based on Wavelet Coefficient Variance.
Xueying Zhang, Jing Bai, Gao‐Feng Zhao · 2006
Abstract- Speech endpoint detection is a key technology for speech recognition. A new speech segmentation algorithm based on the variance of the wavelet coefficients was proposed. Speech signal with noise was decomposed by wavelet into some layers to research the statistic characteristics of every layer wavelet coefficient. The variances of the wavelet coefficients are calculated to form detecting features for endpoint detection. Simulations were made under different signal-to-noise ratios and the results show that this method is efficient to segment noisy speech even at a low signal-to-noise ratio.