Application of Kullback-Leibler Metric to Speech Recognition

Abstract

Article discusses the application of Kullback-Leibler divergence to the recognition of speech signals and suggests three algorithms implementing this divergence criterion: correlation algorithm, spectral algorithm and filter algorithm. Discussion covers an approach to the problem of speech variability and is illustrated with the results of experimental modeling of speech signals. The article gives a number of recommendations on the choice of appropriate model parameters and provides a comparison to some other methods of speech recognition.