Comparative study of voice print Based acoustic features: MFCC and LPCC

Voice is the best biometric feature for investigation and authentication. It has both biological and behavioural features. The acoustic features are related to the voice. The Speaker Recognition System is designed for the automatic authentication of speaker's identity which is truly based on the human's voice. Mel Frequency Cepstrum coefficient (MFCC) and Linear Prediction Cepstrum coefficient (LPCC) are taken in use for feature extraction from the provided voice sample. This paper provides a comparative study of MFCC and LPCC based on the accuracy of results and their working methodology. The results are better if MFCC is used for feature extraction.