Abstract
In this paper, we introduce a fast estimate algorithm for discriminant training of semi-continuous HMM (Hidden Markov Models).
We first present the Frame Discrimination (FD) method proposed in [1] for weight re-estimate. Then, the weight update equation is formulated in the specific framework of semi-continuous models. Finally, we propose an approximated update function which requires a very low level of computational resources.
The first experiments validate this method by comparing our fast discriminant weighting (FDW) to the original one. We observe that, on a digit recognition task, FDW and FD estimate obtain similar results, when our method decreases significantly the computational time.
A second experiment evaluates FDW in Large Vocabulary Continuous Speech Recognition (LVCSR) task. We incorporate semi-continuous FDW models in a Broadcast News (BN) transcription system. Experiments are carried out in the framework of ESTER evaluation campaign ([12]). Results show that in particular context of very compact acoustic models, discriminant weights improve the system performance compared to both a baseline continuous system and a SCHMM trained by MLE algorithm.
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
Woodland, P., Povey, D.: Large scale discriminative training for speech recognition. In: ISCA ITRW Automatic Speech Recognition: Challenges for the Millenium, Paris, pp. 7–16 (2000)
Galliano, S., Geoffrois, E., Mostefa, D., Choukri, K., Bonastre, J.F., Gravier, G.: The ester phase ii evaluation campaign for the rich transcription of French broadcast news. In: Proc. of the European Conf. on Speech Communication and Technology (2005)
Huang, X., Alleva, F., Hon, H.W., Hwang, M.Y., Rosenfeld, R.: The SPHINX-II speech recognition system: an overview. Computer Speech and Language 7(2), 137–148 (1993)
Vaich, T., Cohen, A.: Comparison of continuous-density and semi-continuous hmm in isolated words recognition systems. In: EUROSPEECH 1999, pp. 1515–1518 (1999)
Normandin, Y., Cardin, R., Mori, R.D.: High-performance digit recognition using maximum mutual information. 2, 299–311 (1994)
Aubert, X., Ney, H.: Large vocabulary continuous speech recognition using word graphs. In: Proc. ICASSP 1995, pp. 49–52 (1995)
Lévy, C., Linarès, G., Nocera, P., Bonastre, J.: 7 in Digital Signal Processing for In-Vehicle and Mobile Systems 2. In: Abut, H., Hansen, J.H.L., Takeda, K. (eds.) Embedded mobile phone digit-recognition, Springer Science, Heidelberg (2006)
Lamel, L., Gauvain, J., Eskénazi, M.: BREF, a large vocabulary spoken corpus for French. In: Proceedings of European Conference on Speech Communication and Technology (Eurospeech 1991), Gênes, Italie, pp. 505–508 (1991)
Nocera, P., Fredouille, C., Linarés, G., Matrouf, D., Meignier, S., Bonastre, J.F., Massonié, D., B’ehet, F.: The LIA’s French broadcast news transcription system. In: SWIM: Lectures by Masters in Speech Processing, Maui, Hawaii (2004)
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 2007 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Linarès, G., Lévy, C. (2007). Fast Discriminant Training of Semi-continuous HMM. In: Matoušek, V., Mautner, P. (eds) Text, Speech and Dialogue. TSD 2007. Lecture Notes in Computer Science(), vol 4629. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-74628-7_52
Download citation
DOI: https://doi.org/10.1007/978-3-540-74628-7_52
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-74627-0
Online ISBN: 978-3-540-74628-7
eBook Packages: Computer ScienceComputer Science (R0)