Gender Recognition System for Manipuri Speech Signal by Using MFCC and Power Spectrum

Abstract

In this paper we are introducing a technique for gender classification as male or female. This paper proposed designing of feature extraction of Manipuri speech signal and gender Classification using MFCC and Power Spectrum .Previously there are so many technique for the classification, here we are comparing with all the classifiers and testing the results. The classifiers are used to differentiate the gender as male or female. The database for the speech recognition system is collection of male and female voices. The main features for the gender recognition are the base in the voice, softness, hardness in the voice. We are performing the results by using the Mel frequency cepstrum coefficient (MFCC), power spectrum and Euclidian distance classifier. By seeing the experimental results we can say that the proposed method is giving better results.