Segmentation of continuous speech by HMM and bayesian probability

Abstract

AbstractThis paper describes a new statistical method of segmentation using a hidden Markov model (HMM) and a Bayes' classifier. The main characteristics of this method are the use of feature parameters which are independent of each category in vowels or consonants, and the use of only one HMM which represents all syllable patterns in common. The segmentation strategy is to find the optimal HMM sequence. The optimal/best sequence is found by using the O(n) DP matching based on the Viterbi algorithm. The concatenated number and boundaries of the best HMM sequence are regarded as the segmentation result.