Radial basis functions for speaker identification
Fredrickson SE., Tarassenko L.
Artificial neural networks are applied to the problem of recognising people from their voices. With input vectors consisting of time-normalised cepstral coefficients, radial basis function networks are shown to exceed the speaker classification performance of multilayer perceptrons. With an optimised input vocabulary and the use of multiple observations, error-free results are obtained on a 50 speaker identification task.