Foundations Of Cognitive Science

Winner-Take-All Network

A particular type of artificial neural network, called a winner-take-all network (Feldman & Ballard, 1982), is ideally suited to explain how attention can be automatically drawn to an object or to a distinctive feature (Fukushima, 1986; Gerrissen, 1991; Grossberg, 1980; Koch & Ullman, 1985; LaBerge, Carter, & Brown, 1992; Sandon, 1992).  In a winner-take-all network, an array of processing units is assigned to different objects or to feature locations.  For instance, these processors could be distributed across the preattentive feature maps in feature integration theory (Treisman, 1988; Treisman & Gelade, 1980).    Typically, a processor will have an excitatory connection to itself, and will have inhibitory connections to its neighboring processors.  This pattern of connectivity results in the processor that receives the most distinctive input activating, and at the same time turning its neighbors off.


  1. Feldman, J. A., & Ballard, D. H. (1982). Connectionist models and their properties. Cognitive science, 6, 205-254.
  2. Fukushima, K. (1986). A neural network model for selective attention in visual pattern recognition. Biological Cybernetics, 55, 5-15.
  3. Gerrissen, J. F. (1991). On the network-based emulation of human visual search. Neural Networks, 4, 543-564.
  4. Grossberg, S. (1980). How does the brain build a cognitive code?  Psychological Review, 87, 1-51.
  5. LaBerge, D., Carter, M., & Brown, V. (1992). A network simulation of thalamic circuit operations in selective attention. Neural Computation, 4, 318-331.
  6. Sandon, P. A. (1992). Simulating visual attention. Journal of Cognitive Neuroscience, 2, 213-231.
  7. Treisman, A. M. (1988). Features and objects: The fourteenth Bartlett memorial lecture. Quarterly Journal of Experimental Psychology, 40A, 201-237.
  8. Treisman, A. M., & Gelade, G. (1980). A feature integration theory of attention. Cognitive psychology, 12, 97-136.

(Added March 2011)