Classification of speech files by waveforms

R. Kh. Latypov^a, R. R. Nigmatullin^b, E. L. Stolov<sup>b

a</sup> Institute of Computer Mathematics and Information Technologies, Kazan (Volga Region) Federal University, Kazan, Russia
^b Institute of Computer Mathematics and Information Technologies, Kazan (Volga Region) Federal University, Kazan, Russia

Abstract: The paper presents a new approach to the classification of speech files according to a language being used. Some parameters describing the form of individual signals of a speech file are suggested, such as the position of quantiles in a selected fragment and the parameters of parabolas approximating area under curve. It is proved that distribution of these parameters can be used for automatic differentiation of files containing Tatar and Russian speech.

Keywords: automatic differentiation of languages, form of a file fragment, quantiles, approximation by parabolas.

UDC: 534.4+004.93

Received: 08.09.2014