Dysarthric Speech Classification Using Hierarchical Multilayer Perceptrons and Posterior Rhythmic Features

  • Sid-Ahmed Selouani
  • Habiba Dahmani
  • Riadh Amami
  • Habib Hamam
Part of the Advances in Intelligent and Soft Computing book series (AINSC, volume 87)


In this paper class posterior distributions are combined with a hierarchal structure of multilayer Perceptrons to perform an automatic assessment of dysarthric speech. In addition to the standard Mel-frequency coefficients, this hybrid classifier uses rhythm-based features as input parameters since the preliminary evidence from perceptual experiments show that rhythm troubles may be the common characteristic of various types of dysarthria. The Nemours database of American dysarthric speakers is used throughout experiments. Results show the relevance of rhythm metrics and the effectiveness of the proposed hybrid classifier to discriminate the levels of dysarthria severity.


Cerebral Palsy Speech Signal Gaussian Mixture Model Automatic Assessment Phoneme Recognition 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Arvaniti, A.: A rhythm timing and the timing of rhythm. Phonetica (66), 46–63 (2009)CrossRefGoogle Scholar
  2. 2.
    Enderby, P., Pamela, M.: Frenchay Dysarthria Assessment. College Hill Press (1983)Google Scholar
  3. 3.
    Grabe, E., Low, E.L.: Durational variability in speech and the rhythm class hypothesis. Papers in Laboratory Phonology 7 (2002)Google Scholar
  4. 4.
    Liss, J.M., White, L., Mattys, S.L., Lansford, K., Lotto, A.J., Spitzer, S., Caviness, J.N.: Quantifying speech rhythm abnormalities in the dysarthrias. Journal of Speech Language and Hearing Research (52), 1334–1352 (2009)CrossRefGoogle Scholar
  5. 5.
    Polikoff, J.B., Bunnell, H.T.: The Nemours database of dysarthric speech: A perceptual analysis. In: The XIVth International Congress of Phonetic Sciences (ICPhS), San Francisco (1999)Google Scholar
  6. 6.
    Polur, D., Miller, G.: Investigation of an HMM/ANN hybrid structure in pattern recognition application using cepstral analysis of dysarthric (distorted) speech signals. Medical Engineering & Physics 28(8), 741–748 (2006)CrossRefGoogle Scholar
  7. 7.
    Ramus, F., Nespor, M., Mehler, J.: Correlates of linguistic rhythm in the speech signal. Cognition 73(3), 265–292 (1999)CrossRefGoogle Scholar
  8. 8.
    Rudzicz, F.: Phonological features in discriminative classification of dysarthric speech. In: Proceedings of ICASSP 2009, Taiwan, pp. 4605–4608 (2009)Google Scholar
  9. 9.
    Schwarz, P., Matejka, P., Cernocky, J.: Hierarchical structures of neural networks for phoneme recognition. In: Proceedings of ICASSP 2006, Toulouse, pp. 325–328 (2006)Google Scholar
  10. 10.
    Selouani, S.A., Yakoub, M., O’Shaughnessy, D.: Alternative speech communication system, for persons with severe speech disorders. EURASIP Journal on Advances in Signal Processing (2009), doi:10.1155Google Scholar
  11. 11.
    Tolba, H., Eltorgoman, A.: Towards the improvement of automatic recognition of dysarthric speech. In: IEEE International Conference ICSIT, pp. 277–281 (2009)Google Scholar
  12. 12.
    Tsuji, T., Fukuda, O., Ichinobe, H., Kaneko, M.: A log-linearized Gaussian mixture network and its application to EEG pattern classification. IEEE Transactions on Systems, Man, and Cybernetics 29(1), 60–72 (1999)CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2011

Authors and Affiliations

  • Sid-Ahmed Selouani
    • 1
  • Habiba Dahmani
    • 1
  • Riadh Amami
    • 3
  • Habib Hamam
    • 2
  1. 1.Université de MonctonNew BrunswickCanada
  2. 2.INRS-Université du QuébecMontrealCanada
  3. 3.École ESPRITTunisTunisia

Personalised recommendations