Abstract
A new duration intrinsic model for improved speech recognition by HMM techniques is presented. Assuming an exponentially decaying time dependency of the states loop probability, the duration density can be factorized and a path early pruning theorem demonstrated. As a consequence, computational complexity is greatly reduced with respect to explicit models, whereas recognition performances improve considerably.
This work has been partially founded by ALCATEL-FACE. Only the author is responsible for the ideas and conclusions here reported.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
— J.D.Ferguson, “Variable Duration Models of Speech”, Proc. of Symp. on the application of HMM to Text and Speech, Ferguson jd Ed, Princetown, NJ, 1980
— H.Bourlard, C.J.Welleknes, “Connected Speech Recognition by Phonemic Semi-Markov chains for state occupancy modelling”, Signal Processing III, Theories an applications, I.T. Young et al, Elsevier Sc. Pub. B V (North-Holland)
— L.J.Russel, R.K.Moore, “Explicit modelling of state occupancy in Hidden Markov Models for Automatic Speech Recognition”, Proc.ICASSP-85
— S.E.Levinson, “Continuously Variable Duration Hidden Markov Models for Speech Analysis”, Proc ICASSP-86, p. 1241, Tokyo
- L.R.Rabiner, B.H.Juang, S.E.Levinson, M.M.Sondhi, “Recognition of isolated digits using hidden Markov models with continuous mixture densities”, AT&T Tech.J., Vol.64, pp1211–1234, July-Aug. 1985
- M.Codogno, L.Fissore, “Duration modelling in finite state automata for speech recognition and fast speaker adaptation”, Proc. ICASSP-87, Dallas
- T.H.Crystal, A.S.House, “Characterization and modeling of speech segment duration”, Proc. ICASSP-86, Tokyo
- M.J.Russel, A.E.Cook, “Experimental Evaluation of Durational Modelling Techniques for Automatic Speech Recognition”, Proc ICASSP-87, p. 2376, Dallas
- C.H.Lee, “On the use of some robust modelling techniques for speech recognition”, Computer Speech and Language, 1989, 3, 35–52
- A.Falaschi, “Phonetic recognition by non stationary HMM”, Proc. of FASE Speech 88, Aug. 1988, Edinburgh, UK
- D.Jouvet, J.Monne’, P.Dobois, “A new network-based speaker-independent connected- word recognition system”, Proc. ICASSP-86, Tokyo
- I.S.Gradshtey, I.M.Ryzhik, “Table of integrals series and products” 4th Ed., Acad. Press 1965 — NY, San Francisco, London
- A.Falaschi, “Decodifica acustico-fonetica del messaggio vocale su basi informatico-strutturali e modelli di Markov nascosti”, PhD Thesis, INFO-COM Dpt, La Sapienza Univ. of Rome, Italy
— R.Schwartz, Y.Chow, F.Kubala, “Rapid speaker adaptation using a probabilistic mapping”, Proc of ICASSP-87, Dallas
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1992 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Falaschi, A. (1992). Continuously Variable Transition Probability HMM for Speech Recognition. In: Laface, P., De Mori, R. (eds) Speech Recognition and Understanding. NATO ASI Series, vol 75. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-76626-8_14
Download citation
DOI: https://doi.org/10.1007/978-3-642-76626-8_14
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-76628-2
Online ISBN: 978-3-642-76626-8
eBook Packages: Springer Book Archive