Comparative Analysis of Different Windowing Techniques in MFCC Speaker Recognition (original) (raw)

Speaker recognition is the process of autom atically recognising the speaker on the basis of individual information included in speech waves. The objective of automatic speaker recognition is to extract, characterize and recogni ze the information about speaker identity. Speaker recogni tion technology can be used in many services such as voi ce dialling, banking by telephone, telephone shopping, database access services, information services, voice mail, securit y control for confidential information areas, and remote access t o computers. Feature extraction is an important process in speake r recognition. In this paper Mel Frequency Cepstrum Coef ficients method is used in order to design a text dependent speaker recognition system. Different types of windowing meth ods are used during feature extraction. In this paper, a co mparative analysis of different windowing techniques is done in order to determine the most effective windowing technique for MFCC speaker recognition.