Title :
Minimum cross-entropy adaptation of hidden Markov models
Author :
Afify, Mohamed ; Haton, Jean-Paul
Author_Institution :
Univ. Henri Poincare, Vandeouvre, France
Abstract :
Adaptation techniques that benefit from distribution correlation are important in practical situations having sparse adaptation data. The so called extended MAP (EMAP) algorithm provides an optimal, though expensive, solution. In this article we start from EMAP, and propose an approximate optimisation criterion, based on maximising a set of local densities. We then obtain expressions for these local densities based on the principle of minimum cross-entropy (MCE). The solution to the MCE problem is obtained using an analogy with MAP estimation, and avoids the use of complex numerical procedures, thus resulting in a simple adaptation algorithm. The implementation of the proposed method for the adaptation of HMMs with mixture Gaussian densities is discussed, and its efficiency is evaluated on an alphabet recognition task
Keywords :
Gaussian processes; correlation methods; hidden Markov models; iterative methods; maximum likelihood estimation; minimum entropy methods; optimisation; speech recognition; EMAP algorithm; MAP estimation; adaptation algorithm; alphabet recognition task; approximate optimisation criterion; distribution correlation; efficiency; extended MAP; hidden Markov models; iterative optimisation; minimum cross-entropy adaptation; mixture Gaussian densities; sparse adaptation data; speech recognition systems; Gaussian distribution; Lagrangian functions; Random variables;
Conference_Titel :
Acoustics, Speech and Signal Processing, 1998. Proceedings of the 1998 IEEE International Conference on
Conference_Location :
Seattle, WA
Print_ISBN :
0-7803-4428-6
DOI :
10.1109/ICASSP.1998.674370