Speaker identification using wavelet transform and artificial neural network

Abstract EN

This paper presents an effective method for improving the performance of speaker identification system based on schemes combines the multi resolution properly of the wavelet transform and radial basis function neural networks (RBFNN), evaluated its performance by comparing the results with other method.

The input speech signal is decomposed into L sub band.

To capture the characteristic of the vocal tract, the liner prediction code of each (including the linear predictive code (LPC) for full band) are calculated.

The radial basis function neural network (RBFNN) approach is used for matching purpose.

Experimental results shows that the speaker identification using the methods achieve (combines the wavelet and RBFNN) give (100 %) identification rate and higher identification rate compared with multi band liner predictive code, in this paper used Mat lab program to prove the results..