Statistical and Discriminative Methods for Speech Recognition

Abstract

In this paper, we discuss the issue of speech recognizer training from a broad perspective with root in the classical Bayes decision theory. We differentiate the method of classifier design via distribution estimation and the method of discriminative training based on the fact that in many realistic applications, such as speech recognition, the real signal distribution form is rarely known precisely. We argue that traditional methods relying on distribution estimation are suboptimal when the assumed distribution form is not the true one, and that "optimality" in distribution estimation does not automatically translate into "optimality" in classifier design. We compare the two different methods in the context of hidden Markov modeling for speech recognition. We show the superiority of the discriminative method over the distribution estimation method by citing the results of several key speech recognition experiments. In general, the discriminative method provides a 30-50% reduction in recognition errors.