Skip to main content
Springer Nature Link
Log in

An EM Algorithm to Learn Sequences in the Wavelet Domain

  • Conference paper
MICAI 2007: Advances in Artificial Intelligence (MICAI 2007)

Part of the book series: Lecture Notes in Computer Science ((LNAI,volume 4827))

Included in the following conference series:

Abstract

The wavelet transform has been used for feature extraction in many applications of pattern recognition. However, in general the learning algorithms are not designed taking into account the properties of the features obtained with discrete wavelet transform. In this work we propose a Markovian model to classify sequences of frames in the wavelet domain. The architecture is a composite of an external hidden Markov model in which the observation probabilities are provided by a set of hidden Markov trees. Training algorithms are developed for the composite model using the expectation-maximization framework. We also evaluate a novel delay-invariant representation to improve wavelet feature extraction for classification tasks. The proposed methods can be easily extended to model sequences of images. Here we present phoneme recognition experiments with TIMIT speech corpus. The robustness of the proposed architecture and learning method was tested by reducing the amount of training data to a few patterns. Recognition rates were better than those of hidden Markov models with observation densities based in Gaussian mixtures.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Log in via an institution

Institutional subscriptions

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Similar content being viewed by others

References

  1. Crouse, M., Nowak, R., Baraniuk, R.: Wavelet-based statistical signal processing using hidden Markov models. IEEE Transactions on Signal Processing 46(4), 886–902 (1998)

    Article  MathSciNet  Google Scholar 

  2. Mallat, S.: A Wavelet Tour of signal Processing, 2nd edn. Academic Press, London (1999)

    MATH  Google Scholar 

  3. Fan, G., Xia, X.G.: Improved hidden Markov models in the wavelet-domain. IEEE Transactions on Signal Processing 49(1), 115–120 (2001)

    Article  Google Scholar 

  4. Durand, J.B., Gonçalvès, P., Guédon, Y.: Computational methods for hidden Markov trees. IEEE Transactions on Signal Processing 52(9), 2551–2560 (2004)

    Article  MathSciNet  Google Scholar 

  5. Sebe, N., Cohen, I., Garg, A., Huang, T.: Machine Learning in Computer Vision. Springer, Heidelberg (2005)

    MATH  Google Scholar 

  6. Baldi, P., Brunak, S.: Bioinformatics: The Machine Learning Approach. MIT Press, Cambridge, Masachussets (2001)

    MATH  Google Scholar 

  7. Jelinek, F.: Statistical Methods for Speech Recognition. MIT Press, Cambridge, Masachussets (1999)

    Google Scholar 

  8. Kim, S., Smyth, P.: Segmental Hidden Markov Models with Random Effects for Waveform Modeling. Journal of Machine Learning Research 7, 945–969 (2006)

    MathSciNet  Google Scholar 

  9. Bishop, C.: Neural Networks for Pattern Recognition. Clarendon Press, Oxford (1995)

    Google Scholar 

  10. Rabiner, L., Juang, B.: Fundamentals of Speech Recognition. Prentice-Hall, New Jersey (1993)

    Google Scholar 

  11. Bengio, Y.: Markovian Models for Sequential Data. Neural Computing Surveys 2, 129–162 (1999)

    Google Scholar 

  12. Fine, S., Singer, Y., Tishby, N.: The Hierarchical Hidden Markov Model: Analysis and Applications. Machine Learning 32(1), 41–62 (1998)

    Article  MATH  Google Scholar 

  13. Murphy, K., Paskin, M.: Linear time inference in hierarchical HMMs. In: Dietterich, T., Becker, S., Ghahramani, Z. (eds.) Advances in Neural Information Processing Systems 14, vol. 14, MIT Press, Cambridge (2002)

    Google Scholar 

  14. Willsky, A.: Multiresolution Markov models for signal and image processing. Proceedings of the IEEE 90(8), 1396–1458 (2002)

    Article  Google Scholar 

  15. Dasgupta, N., Runkle, P., Couchman, L., Carin, L.: Dual hidden Markov model for characterizing wavelet coefficients from multi-aspect scattering data. Signal Processing 81(6), 1303–1316 (2001)

    Article  MATH  Google Scholar 

  16. Lu, J., Carin, L.: HMM-based multiresolution image segmentation. IEEE International Conference on Acoustics, Speech and Signal Processing 4, 3357–3360 (2002)

    Google Scholar 

  17. Bengio, S., Bourlard, H., Weber, K.: An EM algorithm for HMMs with emission distributions represented by HMMs. Technical Report IDIAP-RR 11, Martigny, Switzerland (2000)

    Google Scholar 

  18. Weber, K., Ikbal, S., Bengio, S., Bourlard, H.: Robust speech recognition and feature extraction using HMM2. Computer Speech & Language 17(2-3), 195–211 (2003)

    Article  Google Scholar 

  19. Bharadwaj, P., Carin, L.: Infrared-image classification using hidden Markov trees. IEEE Transactions on Pattern Analysis and Machine Intelligence 24(10), 1394–1398 (2002)

    Article  Google Scholar 

  20. Ichir, M., Mohammad-Djafari, A.: Hidden Markov models for wavelet-based blind source separation. IEEE Transactions on Image Processing 15(7), 1887–1899 (2006)

    Article  Google Scholar 

  21. Zue, V., Sneff, S., Glass, J.: Speech database development: TIMIT and beyond. Speech Communication 9(4), 351–356 (1990)

    Article  Google Scholar 

  22. Stevens, K.: Acoustic phonetics. MIT Press, Cambridge (1998)

    Google Scholar 

  23. Mallat, S.: A theory for multiresolution signal decomposition: the wavelet representation. IEEE Transactions on Pattern Analysis and Machine Intelligence 11(7), 674–693 (1989)

    Article  MATH  Google Scholar 

  24. Daubechies, I.: Ten Lectures on Wavelets. In: Number 61 in CBMS-NSF Series in Applied Mathematics, SIAM, Philadelphia (1992)

    Google Scholar 

  25. Evangelista, G.: Pitch-synchronous wavelet representations of speech and music signals. IEEE Transactions on Signal Processing 41(12), 3313–3330 (1993)

    Article  MATH  Google Scholar 

  26. Chan, C.P., Ching, P.C., Leea, T.: Noisy speech recognition using de-noised multiresolution analysis acoustic features. J. Acoust. Soc. Am. 110(5), 2567–2574 (2001)

    Article  Google Scholar 

  27. Farooq, O., Datta, S.: Mel Filter-Like Admissible Wavelet Packet Structure for Speech Recognition. IEEE Signal Processing Letters 8(7) (2001)

    Google Scholar 

Download references

Author information

Authors and Affiliations

  1. Signals and Computational Intelligence Laboratory - CONICET, Department of Informatics, Faculty of Engineering and Water Sciences, National University of Litoral, Ciudad Universitaria, Santa Fe, Argentina

    Diego H. Milone & Leandro E. Di Persia

Authors
  1. Diego H. Milone
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Leandro E. Di Persia
    View author publications

    You can also search for this author in PubMed Google Scholar

Editor information

Alexander Gelbukh Ángel Fernando Kuri Morales

Rights and permissions

Reprints and permissions

Copyright information

© 2007 Springer-Verlag Berlin Heidelberg

About this paper

Cite this paper

Milone, D.H., Di Persia, L.E. (2007). An EM Algorithm to Learn Sequences in the Wavelet Domain. In: Gelbukh, A., Kuri Morales, Á.F. (eds) MICAI 2007: Advances in Artificial Intelligence. MICAI 2007. Lecture Notes in Computer Science(), vol 4827. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-76631-5_49

Download citation

  • DOI: https://doi.org/10.1007/978-3-540-76631-5_49

  • Publisher Name: Springer, Berlin, Heidelberg

  • Print ISBN: 978-3-540-76630-8

  • Online ISBN: 978-3-540-76631-5

  • eBook Packages: Computer ScienceComputer Science (R0)

Publish with us

Policies and ethics