An auditory feature extraction method for robust speaker recognition

Based on simulate human auditory model, the paper proposed an auditory cepstrum coefficient for speaker recognition. This paper analyzed the working mechanism of the human auditory model, and simulated the auditory model of human ear cochlea by gammatone filter banks instead of the traditional triangular filter banks; Based on the nonlinear signal processing capability of human auditory model, exponential compression is used instead of the fixed logarithm compression used in the classic MFCC. Finally, the simulation experiments were conducted based on GMM recognition algorithm. The experimental results show that the proposed auditory cepstrum coefficient has better noise robustness than MFCC and LPCC.