Adaptive Feedback Inhibition Improves Pattern Discrimination Learning

  • Frank Michler
  • Thomas Wachtler
  • Reinhard Eckhorn
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4087)


Neural network models for unsupervised pattern recognition learning are challenged when the difference between the patterns of the training set is small. The standard neural network architecture for pattern recognition learning consists of adaptive forward connections and lateral inhibition, which provides competition between output neurons. We propose an additional adaptive inhibitory feedback mechanism, to emphasize the difference between training patterns and improve learning. We present an implementation of adaptive feedback inhibition for spiking neural network models, based on spike timing dependent plasticity (STDP). When the inhibitory feedback connections are adjusted using an anti-Hebbian learning rule, feedback inhibition suppresses the redundant activity of input units which code the overlap between similar stimuli. We show, that learning speed and pattern discriminatability can be increased by adding this mechanism to the standard architecture.


Lateral Inhibition Feedback Inhibition Output Neuron Lateral Geniculate Nucleus Inhibitory Connection 
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.


  1. Beierlein, M., Gibson, J.R., Connors, B.W.: Two dynamically distinct inhibitory networks in layer 4 of the neocortex. Journal of Neurophysiology 90, 2987–3000 (2003)CrossRefGoogle Scholar
  2. Bi, G., Poo, M.: Synaptic modifications in cultured hippocampal neurons: Dependence on spike timing, synaptic strength, and postsynaptic cell type. The Journal of Neuroscience 18(24), 10464–10472 (1998)Google Scholar
  3. Callaway, E.M.: Local circuits in primary visual cortex of the macaque monkey. Annual Review of Neuroscience 21, 47–74 (1998)CrossRefGoogle Scholar
  4. Eckhorn, R.: Neural mechanisms of scene segmentation: Recordins from the visual cortex suggest basic circuits for linking field models. IEEE Transactions on Neural Networks 10(3), 464–479 (1999)CrossRefGoogle Scholar
  5. Eckhorn, R., Bruns, A., Gabriel, A., Al-Shaikhli, B., Saam, M.: Different types of signal coupling in the visual cortex related to neural mechanisms of associative processing and perception. IEEE Transactions on Neural Networks 15(5), 1039–1052 (2004)CrossRefGoogle Scholar
  6. Fukushima, K.: Cognitron: A self-organizing multilayered neural network. Biological Cybernetics 20, 121–136 (1975)CrossRefGoogle Scholar
  7. Földiák, P.: Forming sparse representations by local anti-hebbian learning. Biological Cybernetics 64, 165–170 (1990)CrossRefGoogle Scholar
  8. Grossberg, S.: Linking the laminar circuits of visual cortex to visual perception: Development, grouping and attention. Neuroscience and Biobeavioral Revies 25, 513–526 (2001)CrossRefGoogle Scholar
  9. Izhikevich, E.M.: Simple model of spiking neurons. IEEE Transactions on Neural Networks 14(6), 1569–1572 (2003)CrossRefMathSciNetGoogle Scholar
  10. Izhikevich, E.M.: Which model to use for cortical spiking neurons? IEEE Transactions on Neural Networks 15(5), 1063–1070 (2004)CrossRefGoogle Scholar
  11. Miyake, S., Fukushima, K.: A neural network model for the mechanism of feature-extraction. A self-organizing network with feedback inhibition. Biological Cybernetics 50, 377–384 (1984)MATHCrossRefGoogle Scholar
  12. Rao, R.P.N., Ballard, D.H.: Dynamic model of visual recognition predicts neural response properties in the visual cortex. Neural Computation 9, 721–763 (1997)CrossRefGoogle Scholar
  13. Royer, S., Paré, D.: Conservation of total synaptic weight through balanced synaptic depression and potentiation. Nature 422, 518–522 (2003)CrossRefGoogle Scholar
  14. Spratling, M.W.: Pre-synaptic lateral inhibition provides a better arcitecture for self-organizing neural networks. Network: Computation in Neural Systems 10, 285–301 (1999)MATHCrossRefGoogle Scholar
  15. Spratling, M.W., Johnson, M.H.: Pre-integration lateral inhibition enhances unsupervised learning. Neural Computation 14(9), 2157–2179 (2002)MATHCrossRefGoogle Scholar
  16. van Ooyen, A., Nienhuis, B.: Pattern recognition in the neocognitron is improved by neuronal adaptation. Biological Cybernetics 70, 47–53 (1993)MATHCrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • Frank Michler
    • 1
  • Thomas Wachtler
    • 1
  • Reinhard Eckhorn
    • 1
  1. 1.AppliedPhysics/NeuroPhysics Group, Department of PhysicsPhilipps-University MarburgMarburgGermany

Personalised recommendations