Malay continuous speech recognition using fast HMM match algorithm

Chee Ming Ting, Sh Hussain Salleh, A. K. Ariff

Research output: Chapter in Book/Report/Conference proceedingConference PaperResearchpeer-review

7 Citations (Scopus)


This paper describes the implementation of fast hidden Markov model (HMM) match algorithm in a phoneme-based Malay continuous speech recognition system. The decoding algorithm decouples the computations of state-likelihoods of phone HMM's from the main search which is bounded by syntactical and lexical constraints. This avoids the redundant state-likelihood computations of identical phone HMM's for different word models in the tightly integrated search and thus substantially reduce the decoding time. The algorithm is implemented in the framework of one pass dynamic programming search. For a 541-word task, the fast HMM match reduces the real-time factor (RTF) by a factor of 31.8 from 286.45×RT to 9.02×RT, compared to without decoupling. The word accuracy is maintained at 91.6% without loss for a test set perplexity of 15.45 in speaker-dependent mode.

Original languageEnglish
Title of host publication2009 4th IEEE Conference on Industrial Electronics and Applications, ICIEA 2009
Number of pages4
Publication statusPublished - 2009
Externally publishedYes
EventIEEE Conference on Industrial Electronics and Applications 2009 - Xi'an, China
Duration: 25 May 200927 May 2009
Conference number: 4th (Proceedings)


ConferenceIEEE Conference on Industrial Electronics and Applications 2009
Abbreviated titleICIEA 2009
Internet address


  • Acoustic match
  • Continuous speech recognition
  • Hidden markov model

Cite this