마진이 강화된 최대 상호 정보 방법을 이용한 은닉 마르코프 모델의 매개변수 추정

Abstract

We considered a discriminative training algorithm to estimate continuous-density hidden Markov model for speech recognition. The proposed algorithm, called margin-enhanced maximum mutual information (MEMMI), is to maximize the weighted sum of the maximum mutual information objective function and the large margin objective function. The MEMMI leads to a simple objective function that can be optimized easily by a gradient ascent algorithm maintaining a probabilistic model. Experimental results show that the recognition accuracy of the MEMMI is better than other discriminative training criteria on the TIDIGITS database.