A Study on Speech Recognition using New State- Clustering Algorithm of HM-Net with Korean Phonological Rules

摘要

Recent large vocabulary continuous speech recognition (LVCSR) systems are based on state-clustered hidden Markov models (HMMs). By Successive State Splitting (SSS) algorithm, the Hidden Markov Network (HM-Net), which is an efficient representation of phoneme-context-dependent HMMs, can be generated automatically. SSS is a powerful technique to design topologies of tied-state HMMs, but it doesn't treat unknown contexts in the training phoneme contexts environments adequately. In addition it has some problem in the procedure of the contextual domain. In this paper, we adopt a new state-clustering algorithm of SSS, called Phonetic Decision Tree-based SSS (PDT-SSS), which includes contexts splits based on the Korean phonological rules. This method combines advantages of both the decision tree clustering and SSS, and can generate highly accurate HM-Nets that can express any contexts. Through the Korean phoneme and word recognition experiments, we proved that the new state-clustering algorithm produce belter phoneme and word recognition accuracy than the conventional HMMs.