ICANN ’93 pp 43-46 | Cite as

Neurobiological Modelling and Structured Neural Networks

  • B. Brückner
  • W. Zander


We are presenting a coupled modelling concept which is suitable to reproduce sensory systems in a plausible way. The idea is to connect a biological relevant Neural Network as a filter with a structured formal Neural Network to classify the filtered data. Because the biological filter is very flexible and self organized it has the capacity to make segmentation and feature extraction of input data like a sensory pathway. So, temporal processes can be represented in a sequence of prefiltered segments. The produced time sequences are suited for the analysis by a multi-stage construction of the classification network, which has a modified Hypermap structure. The presented system is succesfully used for speech recognition of different subjects.


Sensory Node Feature Extraction Input Vector Speech Recognition Speech Signal 
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]
    Gordon M. Shepherd. Neurobiology. Qxford University Press, New York, Oxford, 1988.Google Scholar
  2. [2]
    H. Handwerker. Allgemeine Sinnesphysiologie. In R. F. Schmidt and G. Thews, editors, Physiologie des Menschen, pages 186–206. Springer-Verlag, Berlin, 1987.CrossRefGoogle Scholar
  3. [3]
    W. D. Keidel. Biokybernetik des Menschen. Wissenschaftliche Buchgesellschaft, Darmstadt, 1989.Google Scholar
  4. [4]
    A. Richter, W. Zander, and E. Körner. Using parametric controlled structured networks to approach neural networks to neurosciences. In Kohonen et al. [13], pages 1451–1454.Google Scholar
  5. [5]
    W. Zander, B. Brückner, and G. Brankatschk. Simulation of biological relevant neural networks. In Contributions to the 21th Göttingen Neurobiology Conference. Georg-Thieme-Verlag, 1993. in press.Google Scholar
  6. [6]
    Bernd Brückner, Marcella Franz, and Andreas Richter. A modified hypermap architecture … In I. Aleksander and J. Taylor, editors, Artificial Neural Networks, 2, pages 1167–1170, Amsterdam, 1992. Elsevier Science Publishers.Google Scholar
  7. [7]
    Bernd Brückner and Wiebke Zander. Classification of speech using a modified hypermap architecture. submitted to WCNN’93.Google Scholar
  8. [8]
    B. Brückner, M. Franz, and A. Richter. Analyse und klassifizierung von eeg-daten mittels strukturierter neuronaler netze. In Tagungsband 36.Internat, wiss. Koll. der TE Ilmenau, pages 694–699, Ilmenau, 1991.Google Scholar
  9. [9]
    H. Ritter, T. Martinetz, and K. Schulten. Neuronale Netze. Addison-Wesley, Bonn, 1990.Google Scholar
  10. [10]
    Teuvo Kohonen. The hypermap architecture. In Kohonen et al. [13], pages 1357–1360.Google Scholar
  11. [11]
    Jari Kangas. Time-dependent self-organizing maps for speech recognition. In Kohonen et al. [13], pages 1591–1594.Google Scholar
  12. [12]
    T. Kohonen. The self-organizing map. Proceedings of the IEEE, 78(9): 1464–1480, 1990.Google Scholar
  13. [13]
    T. Kohonen, K. Mäkisara, O. Simula, and J. Kangas, editors. Artificial Neural Networks, Helsinki, 1991. Elsevier Science Publishers.Google Scholar

Copyright information

© Springer-Verlag London Limited 1993

Authors and Affiliations

  • B. Brückner
    • 1
  • W. Zander
    • 1
  1. 1.Department acoustics, learning, speechInstitute of Neurobiology MagdeburgMagdeburgGermany

Personalised recommendations