Training HMMs using a minimum recognition error approach

The authors present results on the evaluation of an HMM (hidden Markov model) training procedure which aims explicitly at reducing the recognition error and increasing the discrimination between competing phonetic classes. They propose a function of the model parameters and the training data which can be interpreted as a temporal integration of a `frame recognition error´. A heuristic iterative algorithm is proposed in which the parameters of the probability distributions associated with each state of the HMM are modified with the objective of reducing the value of this error function. Experimental evaluation of the proposed training algorithm in a task of acoustic-phonetic decoding showed improved recognition performance relative to the values obtained with maximum likelihood training