Shaxsni nutqi orqali identifikatsiyalash algoritmlarini amalga oshirish
Annotatsiya
Human speech is a complex biosignal that embodies the unique biometric characteristics of a person related to the personality and state of the speaker. Speaker identification by voice is divided into text dependent and textindependent. The implementation of this system requires many complex mathematical and algorithmic solutions. In particular, the implementation of such systems in hardware and software requires minimizing the requirements for computing resources.This article examines the stages of real time implementation of the speaker's personality identification system based on the spoken word. the voice of the speaker is recorded for use as a key word for identification. After passing through the preprocessing stages, the features are separated, trained and stored in the system. Any incoming speech of the unknown speaker into the system through the microphone is received as a keyword based on the vector quantization method is compared with the biometric parameters of the speakers stored in advance. The accuracy of the speaker identification algorithm using the selected method was 96%.