The Use of Discrete Meyer Wavelet for Speech Segmentation

This work describes the segmentation of oral speech in the Yakut language. The phonetic structure of the Yakut language is described. For segmentation of sound recording, wavelet analysis is used. The choice of a discrete Meyer wavelet is based. The wavelet library of the Scilab open source software package is described. The results of calculations in the Scilab package are presented. The automatic segmentation of oral speech in the Yakut language is investigated. An example of sound segmentation is given. Reasons for reducing accuracy and complexity in speech segmentation are identified.