A novel approach for speaker recognition through wavelet analysis by using artificial neural networks