Loosely coupled HMMs for ASR

  title={Loosely coupled HMMs for ASR},
  author={Harriet J. Nock and Steve J. Young},
Hidden Markov Models (HMMs) have been successful for modelling the dynamics of carefully dictated speech, but their performance degrades severely when used to model conversational speech. This paper presents a preliminary feasibility study of an alternative class of models: loosely coupled HMMs. Since speech is produced by a system of loosely coupled articulators, stochastic models explicitly representing this parallelism may have advantages for automatic speech recognition (ASR), particularly… CONTINUE READING
Highly Cited
This paper has 35 citations. REVIEW CITATIONS


Publications referenced by this paper.
Showing 1-10 of 12 references

and P Woodland

  • S Young, J Jansen, J Odell, D Ollason
  • The HTK Book (Version 2.0). ECRL,
  • 1995
Highly Influential
7 Excerpts

Loosely Coupled HMMs For ASR: A Preliminary Study

  • HJ Nock, SJ Young
  • Technical Report CUED/FINFENG/TR386, CUED,
  • 2000
Highly Influential
4 Excerpts

A Multi-Band Approach to Automatic Speech Recognition

  • N Mirghafori
  • PhD thesis, ICSI, UC Berkeley,
  • 1999
1 Excerpt

and MI Jordan

  • Z Ghahraman
  • Factorial Hidden Markov Models.Machine Learning…
  • 1997

The ISOLET spoken letter database

  • R Cole, Y Muthusamy, M Fanty
  • Technical Report CSE 90-004, OGI,
  • 1990

Similar Papers

Loading similar papers…