Speaker’s Voice Recognition Methods in High-Level Interference Conditions

The statistical and combined methods of identifying the speaker's voice based on the pitch frequency in high-level interference conditions have been developed. Determination of informative spectral components of a speaker's useful audio signal in conditions of text-dependent and text-independent identification has been conducted. The estimation of effectiveness of implementing the developed methods through the use of statistical decision-making procedures and classifying the parameters of feature space of identification objects has been formed.