Wavelet Based Noise Robust Features for Speaker Recognition

Extraction and selection of the best parametric representation of acoustic signal is the mostimportant task in designing any speaker recognition system. A wide range of possibilities existsfor parametrically representing the speech signal such as Linear Prediction Coding (LPC) ,Melfrequency Cepstrum coefficients (MFCC) and others. MFCC are currently the most popularchoice for any speaker recognition system, though one of the shortcomings of MFCC is that thesignal is assumed to be stationary within the given time frame and is therefore unable to analyzethe non-stationary signal. Therefore it is not suitable for noisy speech signals. To overcome thisproblem several researchers used different types of AM-FM modulation/demodulation techniquesfor extracting features from speech signal. In some approaches it is proposed to use the waveletfilterbanks for extracting the features. In this paper a technique for extracting the features bycombining the above mentioned approaches is proposed. Features are extracted from theenvelope of the signal and then passed through wavelet filterbank. It is found that the proposedmethod outperforms the existing feature extraction techniques.