This paper proposes a new paradigm to compensate for mismatch condition in speech recognition. A two-step Viterbi decoding based on reinforcement learning is described. The idea is to strength or weaken HMM's by using Bayes-based confidence measure ( BBCM) and distances between models. If HMM's in the N-best list show a low BBCM, the second Viterbi decoding will prioritize the search on neighboring models according to their distances to the N-best HMM's. As shown here, a reduction of 6% in WER is achieved in a task which results difficult for standard MAP and MLLR adaptation.
Cite as: Molina, C., Yoma, N.B., Huenupán, F., Garreton, C. (2007) Unsupervised re-scoring of observation probability in viterbi based on reinforcement learning by using confidence measure and HMM neighborhood. Proc. Interspeech 2007, 1733-1736, doi: 10.21437/Interspeech.2007-486
@inproceedings{molina07_interspeech,
author={Carlos Molina and Nestor Becerra Yoma and Fernando Huenupán and Claudio Garreton},
title={{Unsupervised re-scoring of observation probability in viterbi based on reinforcement learning by using confidence measure and HMM neighborhood}},
year=2007,
booktitle={Proc. Interspeech 2007},
pages={1733--1736},
doi={10.21437/Interspeech.2007-486}
}