Speaker identification using Bessel function representation and a back-propagation neural network

This paper proposes a method of identifying speakers using a Bessel function expansion of speech and a backpropagation neural network. Selected coefficients in the first order Fourier-Bessel expansion series of speech are used as input to the neural network. Based on initial testing on a small set of speakers, the results show an increase in recognition accuracy as the number of coefficients is is increased.