Comparison Between Two Spatio-Temporal Organization Maps for Speech Recognition

  • Zouhour Neji Ben Salem
  • Laurent Bougrain
  • Frédéric Alexandre
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4087)


In this paper, we compare two models biologically inspired and gathering spatio-temporal data coding, representation and processing. These models are based on Self-Organizing Map (SOM) yielding to a Spatio-Temporel Organization Map (STOM). More precisely, the map is trained using two different spatio-temporal algorithms taking their roots in biological researches: The ST-Kohonen and the Time-Organized Map (TOM). These algorithms use two kinds of spatio-temporal data coding. The first one is based on the domain of complex numbers, while the second is based on the ISI (Inter Spike Interval). STOM is experimented in the field of speech recognition in order to evaluate its performance for such time variable application and to prove that biological models are capable of giving good results as stochastic and hybrid ones.


Speech Recognition Speech Signal Automatic Speech Recognition Inter Spike Interval Digit Recognition 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Agmon-Snir, H., Segev, I.: Signal Delay and iInput Synchronization in Passive Dendritic Structures. Journal of Neurophysiology 70(5) (1973)Google Scholar
  2. 2.
    Cariani, P.: As If Time Really Mattered: Temporal Strategies for Neural Coding of Sensory Information. CC-AI 12(1-2) (1995)Google Scholar
  3. 3.
    Spengler, F., Hilger, T., Wang, X., Merzenich, M.: Learning induced formation of cortical populations involved in tactile object recognition. Social Neurosciences 22, 105–110 (1999)Google Scholar
  4. 4.
    Wilson, H., Cowan, J.: A mathematical theory of the functional dynamics of cortical and thalamic nervous tissue. Biology Cybernetic 13, 55–80 (1973)Google Scholar
  5. 5.
    Amari, S.: Topographic organization of nerve fields. Bull Math Biology 42, 339–364 (1980)MATHMathSciNetGoogle Scholar
  6. 6.
    Szentagothai, J.: The module concept in cerebral cortex architecture. Brain research, 47–496 (1995)Google Scholar
  7. 7.
    Vinh ho, H.: Un reseau de neurons a decharge pour la reconnaissance des processus spatio-temporels. PhD thesis, Genie Electric Department, Monreal University (1992)Google Scholar
  8. 8.
    Wiemer, J., Spengler, F., Joublin, F., Wacquant, S.: Learning cortical topography from spatiotemporal stimuli. Biology cybernetic 82, 173–187 (2000)CrossRefGoogle Scholar
  9. 9.
    Casti, A.R.R., Omurtag, A., Sornborger, A., Aplan, E., Knight, B., Victor, J., Sirovich, L.: A population study of integrate and fire or burst neuron. Neural Computation 14(5) (2002)Google Scholar
  10. 10.
    Laurence, S., Tsoi, A.C., Back, A.D.: The gamma MLP for speech phoneme recognition. Advances in Neural Information Processing System 8, 785–791 (1996)Google Scholar
  11. 11.
    Wiemer, J.C.: The Time-Organized Map (TOM) algorithm: extending the self-organizing map (SOM) to spatiotemporal signals. Neural Networks 15 (2003)Google Scholar
  12. 12.
    Vaucher, G.: A la recherche d’une algerbre neuronale spatio-temporal. P.hD thesis. Nancy University (1996)Google Scholar
  13. 13.
    Mozayyani, N., Alanou, V., Derfus, J., Vaucher, G.: A spatio-temporal data coding applied to kohonen maps. In: Inter. conf. on Artificial Neural Network, pp. 75–79 (1995)Google Scholar
  14. 14.
    Baig, A.B.: Une approche methodologique de l’utilisation des STAN applique a la reconnaissance visuelle de la parole. PhD thesis, Suplec, campus universitaire de rennes (2000)Google Scholar
  15. 15.
    Vaucher, G.: A Complex-Valued spiking machine. In: Kaynak, O., Alpaydın, E., Oja, E., Xu, L. (eds.) ICANN 2003 and ICONIP 2003. LNCS, vol. 2714, pp. 967–976. Springer, Heidelberg (2003)CrossRefGoogle Scholar
  16. 16.
    Thorpe, S.: Spiking arrival times: A highly efficient coding scheme for neural networks. In: Parallel Processing in Neural System, Elseiver Press, Amsterdam (1990)Google Scholar
  17. 17.
    Rall, W.: Core conductor theory and cable properties. In: Handbook of physiology: the nervous system, Americain physiology society (1977)Google Scholar
  18. 18.
    Calliope. La parole et son traitement automatique. Masson, Paris, Milan, Barcelone (1989)Google Scholar
  19. 19.
    Durand, S.: Learning speech as acoustic sequences with the unsupervised model TOM. In: NEURAP, 8th international conference on neural networks and their applications, Marseille french (1995)Google Scholar
  20. 20.
    Béroulle, D.: Un modèle de mémoire adaptative, dynamique et associative, pour le traitement automatique de la parole. Thèse de l’université de Paris 11 (1985)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • Zouhour Neji Ben Salem
    • 1
  • Laurent Bougrain
    • 2
  • Frédéric Alexandre
    • 2
  1. 1.AI Unit, CRISTAL LaratoryNational School of Computer SciencesTunisia
  2. 2.Cortex Team, LORIA LaboratoryNancyFrance

Personalised recommendations