Incremental training of CDHMMs using bayesian learning
نویسندگان
چکیده
The Bayesian Learning approach (MAP Maximum A Posteriori) can be used for the incremental training of Continuous Density Hidden Markov Models (CDHMM), performed through speech data collected in real applications. The effectiveness of MAP is heavily conditioned by the correct balance between the a-priori knowledge and the field training data. In this paper we propose and evaluate several optimization methods of the MAP combination function, based either on maximum likelihood (ML) and heuristics criteria. To adjust the relevance of the a-priori knowledge we use the exponential forgetting technique into the MAP framework. We present several tests that compare the error rate reduction as a function of the selected optimization method and of the size of adaptation data.
منابع مشابه
I Summary of Best Ti and Td Vq Performance Speaker Recognition Using Hidden Markov Models, Dynamic Time Warping and Vector Quantisation
1 Illustration of the segmentation of the database collected over a period of three months into training and 3 %Error against total number of mixtures for TI ergodic CDHMMs (10 version training) 7 %Error against the number of training versions for a TI 32 element VQ, and 32 mixture single state CDHMM 11 8 %Error against the number of training versions for TD DTW, 8 element VQ and 1 mixture 8 st...
متن کاملIncremental Naïve Bayesian Learning Algorithm based on Classification Contribution Degree
In order to improve the ability of gradual learning on the training set gotten in batches of Naive Bayesian classifier, an incremental Naïve Bayesian learning algorithm is improved with the research on the existing incremental Naïve Bayesian learning algorithms. Aiming at the problems with the existing incremental amending sample selection strategy, the paper introduced the concept of sample Cl...
متن کاملOn-line adaptive learning of CDHMM parameters based on multiple-stream prior evolution and posterior pooling
Based on the concept of multiple-stream prior evolution and posterior pooling, we propose a new incremental adaptive Bayesian learning framework for e cient on-line adaptation of the continuous density hidden Markov model (CDHMM) parameters. As a rst step, we apply the a ne transformations to the mean vectors of CDHMMs to control the evolution of their prior distribution. This new stream of pri...
متن کاملTraining of context-dependent subspace distribution clustering hidden Markov model
Training of continuous density hidden Markov models (CDHMMs) is usually time-consuming and tedious due to the large number of model parameters involved. Recently we proposed a new derivative of CDHMM, the subspace distribution clustering hidden Markov model (SDCHMM) which tie CDHMMs at the ner level of subspace distributions, resulting in many fewer model parameters. An SDCHMM training algorith...
متن کاملBayesian adaptation of speech recognizers to field speech data
This work studies a Bayesian (or Maximum A Posteriori MAP) approach to the adaptation of Continuous Density Hidden Markov Models (CDHMMs) to a specific condition of a speech recognition application. In order to improve the model robustness, CDHMMs formerly trained from laboratory data are then adapted using context dependent field utterances. Two specific problems have to be faced when using th...
متن کامل