Unsupervised re-scoring of observation probability in viterbi based on reinforcement learning by using confidence measure and HMM neighborhood

Molina, Carlos; Yoma, Nestor Becerra; Huenupán, Fernando; Garreton, Claudio

doi:10.21437/Interspeech.2007-486

Unsupervised re-scoring of observation probability in viterbi based on reinforcement learning by using confidence measure and HMM neighborhood

Carlos Molina, Nestor Becerra Yoma, Fernando Huenupán, Claudio Garreton

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.