Rights Creative Commons: Attribution 3.0 Hong Kong License IMPROVING VITERBI BAYESIAN PREDICTIVE CLASSIFICATION VIA SEQUENTIAL BAYESIAN LEARNING IN ROBUST SPEECH RECOGNITION

In this paper, we extend our previously proposed Viterbi Bayesian predictive classification (VBPC) algorithm to ac-commodate a new class of prior probability density function (pdf) for continuous density hidden Markov model (CDHM-M) based robust speech recognition. The initial prior pdf of CDHMM is assumed to be a finite mixture of natural con-jugate prior pdf’s of its complete-data density. With the new observation data, the true posterior pdf is approximat-ed by the same type of finite mixture pdf’s which retain the required most significant terms in the true posterior den-sity according to their contribution to the corresponding predictive density. Then the updated mixture pdf is used to improve the VBPC performance. The experimental re-sults on a speaker-independent recognition task of isolated Japanese digits confirm the viability and the usefulness of the proposed technique. 1