Main methods of identifying a speaker through speech

Abstract

This paper provides an overview of methods used for voice recognition. Special attention is given to the robust structure of recognition systems. Additionally, a brief description of the most widely used feature extraction methods (such as MFCC and LPCC) and classification techniques (vector quantization method, Gaussian mixture model, support vector method) is presented. Methods for evaluating recognition systems and approaches for presenting the results of such evaluations are discussed.