A Comparison of MFCC and LPCC with Deep Learning for Speaker Recognition

Abstract

The biological information includes a fingerprint, an iris, a face, a vein, a voice. Among them, since the voice is not touched directly, the psychological burden on the user at the time of input is small as compared with other biological information. In addition, it is possible to input voice easily. Speaker recognition refers to automatically determining whose voice by the characteristics of individuality included in the voice. In order to perform individual authentication, it is necessary to extract the characteristics of the speaker from the voice data. One of the methods is Mel frequency cepstral coefficient (MFCC) and other is linear prediction cepstral coefficients (LPCC). In this paper, the speaker recognition is performed using MFCC and LPCC features by using a neural network with deep learning, we will evaluate features of MFCC and LPCC on speaker recognition.