Voice signal segmentation using wavelet transforms

This article provides a voice channel segmentation algorithm using the multiresolution analysis and wavelet transform, which features both within the stationary sites significant role for the analysis of language play signal spectral characteristics that are determined by the transfer function of the vocal tract, and is changed in the process of articulation. I.e., speech is characterized by nonlinear fluctuations of different size, which are unique to the various speakers. Proposed in this paper algorithm modification for multiresolution analysis allows you to find borders labels of interphoneme transitions for the sounds of different speakers.