Duration modeling with expanded HMM applied to speech recognition
Document typeConference lecture
PublisherH. TIMOTHY BRUMMELL, WILLIAM IDSARDI CITATION DELAWARE, NEW CASTLE, DELAWARE
Rights accessOpen Access
The occupancy of the HMM states is modeled by means of a Markov chain. A linear estimator is introduced to compute the probabilities of the Markov chain. The distribution function (DF) represents accurately the observed data. Representing the DF as a Markov chain allows the use of standard HMM recognizers. The increase of complexity is negligible in training and strongly limited during recognition. Experiments performed on acoustic-phonetic decoding shows how the phone recognition rate increases from 60.6 to 61.1. Furthermore, on a task of database inquires, where phones are used as subword units, the correct word rate increases from 88.2 to 88.4.
CitationBonafonte, A., Vidal, J., Nogueiras, A. Duration modeling with expanded HMM applied to speech recognition. A: International Conference on Spoken Language. "Fourth international conference on spoken language, 1996, ICSLP 96: proceedings". Philadelphia, PA: H. TIMOTHY BRUMMELL, WILLIAM IDSARDI CITATION DELAWARE, NEW CASTLE, DELAWARE, 1996, p. 1097-1100.