Abstract
Motivated by applications to automatic machine recognition of speech, nongaussian parametric density estimation for classification of acoustic feature vectors, which are known to have very high dimensionality, is studied in a maximum likelihood framework. We use EM type algorithms for the estimation of parameters for a mixture model of nongaussian densities. Our experience with these techniques in the context of large vocabulary continuous speech recognition is reported. Experimental results tend to indicate that nongaussian mixture component densities model speech data more effectively from this point of view. Comments are made on the convergence of the iterative estimation algorithms developed.
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
Jelenik, F. (1997). Statistical Methods for Speech Recognition. MIT Press.
Abramowitz, M. and Stegun, I. (1972). Handbook of Mathematical Functions. Dover Publications, New York, Ninth Dover printing.
Redner, R. and Walker, H. (1984). Mixture densities, maximum likelihood and the EM algorithm. SIAM Review, Vol.26, No.2.
Titterington, D. M., Smith, A. F. M. and Markov, U. E. (1985). Statistical analysis of finite Mixture Distributions. Wiley Interscience, New York.
Scott, D. W. (1997). Multivariate Density Estimation. Wiley Interscience.
Thompson, J. R. and Tapia, R. A. (1997). Nonparametric Function Estimation, modelling and simulation, SIAM Publications.
Hartigan, J. (1975). Clustering Algorithms. John Wiley & Sons.
Jain, A. and Dubes, R. (1988). Algorithms for Clustering Data, Prentice Hall.
Bishop, C. M. (1997). Neural Networks for Pattern Recognition, Cambridge University Press.
Devroye, L. and Gyorfi, L. (1984). Non-Parametric Density Estimation: The L1 view. Wiley Interscience.
Fukunaga, K. (1990). Statistical Pattern Recognition, second edition, Academic Press.
Horn, R. A. and Johnson, C. R. (1985). Matrix Analysis, Cambridge University Press.
Vapnik, V. N. (1995). The Nature of Statistical Learning Theory, Springer Verlag.
Baum, L. E., Petrie, T., Soules, G., Weiss, N. (1970). A maximization technique occurring in the statistical analysis of probabilistic functions of Markov chains. The Annals of Math. Stat., Vol.41, No.1, pp.164–171.
Dempster, A. P., Laird, N. M. and Baum, D. B. (1977). Maximum likelihood from incomplete data via the EM algorithm. Journal of Royal Statistical Soc., Ser. B, Vol. 39, pp. 1–38.
Marroquin, F. and Girosi, F. (1993). Some extensions of the K-means algorithm for image segmentation and pattern classifcation, MIT Artificial Intelligence Lab. A. I. Memorandum no. 1390.
Ney, H., Noll, A. (1988). Phoneme modelling using continuous mixture densities. Proceedings of IEEE Int. Conf. on Acosutics Speech and Signal Processing, pp.437–440.
Bahl, L. R., Desouza, P. V., Gopalkrishnan, P. S., Picheny, M. A. (1993). Context dependent vector quantization for continuous speech recognition. Proceedings of IEEE Int. Conf. on Acosutics Speech and Signal Processing, pp.632–635.
Breiman, L. (1983). Classification and Regression Trees, Wadsworth International, Belmont, California.
ARPA Wall Street Journal data, Available from Linguistics Data Corporation.
Geman, S. (1998). Three lectures on image understanding. The Center For Imaging Science, Washington State University, video tape, Sept. 10-12, 1997. Also, David Mumford, Pattern theory, Lecture in Directors Series, IBM Yorktown Heights.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1998 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Basu, S., Micchelli, C.A. (1998). Parametric Density Estimation for the Classification of Acoustic Feature Vectors in Speech Recognition. In: Suykens, J.A.K., Vandewalle, J. (eds) Nonlinear Modeling. Springer, Boston, MA. https://doi.org/10.1007/978-1-4615-5703-6_4
Download citation
DOI: https://doi.org/10.1007/978-1-4615-5703-6_4
Publisher Name: Springer, Boston, MA
Print ISBN: 978-1-4613-7611-8
Online ISBN: 978-1-4615-5703-6
eBook Packages: Springer Book Archive