Speaker Identification Using Combined MFCC and Phase Information

Abstract: Speaker recognition is the identification of the person who is speaking by the characteristics of their voices. To improve the performance of speaker recognition systems, an effective and robust method is proposed to extract speech features, capable of operating in noisy environment. For capturing the characteristics of the signal, the Mel-Frequency Cepstral Coefficients (MFCC) are calculated. Gaussian Mixture Models (GMMs) are used for the recognition stage as they give better recognition for the speakers’ features. In conventional speaker recognition methods based on MFCC, phase information has been ignored. The proposed method integrated the phase information with MFCC on the speaker recognition method. Comparison of the proposed approach with the MFCCs conventional feature extraction method shows that the proposed method improves the recognition rate.

Keywords: Speaker Identification, Mel-Frequency Cepstral Coefficients (MFCCs), Phase Information, GMM, EM Algorithm.