METHOD OF SEGMENTATION OF SPEECH SIGNAL INTO PHONEMS

Abstract

This article discusses the advantage of the wavelet transform over traditional signal processing methods. A method of phonemic segmentation of a speech signal is considered by analyzing the change in the energy of a speech signal at each level of the wavelet transform. A step-by-step algorithm of the proposed method is described. Based on the proposed algorithm, the result of the segmentation process into phonemic units is given.