VQ Based Comparative Analysis of MFCC and LPC Speaker Recognition System

Language is the foremost part of communication and speech is the main component of language. Speaker recognizers extract parameters from speech and characterize the speech signal for recognition. Speaker recognition is a significant part of digital signal processing but the speaker recognition system is greatly affected by noise. Noise degrades the performance of the system. With the rapid advancement of speech processing technology, the speaker recognition system has also improved to a great extent. Different feature extraction techniques have been proposed. This paper investigates the performance of the Mel Frequency Cepstral Coefficient (MFCC) and Linear Predictive coefficients (LPC) in speaker recognition using Vector Quantization (VQ) and proposed some optimization techniques for the best recognition system.