May 2004
Volume 45, Issue 13
ARVO Annual Meeting Abstract  |   May 2004
A computational model of the motion aftereffect
Author Affiliations & Notes
  • L.R. Kozak
    IBILI Center of Ophthalmology, Coimbra University, Coimbra, Portugal
    Institute for Psychology, Hungarian Academy of Sciences, Budapest, Hungary
  • M. Castelo–Branco
    IBILI Center of Ophthalmology, Coimbra University, Coimbra, Portugal
  • G. Karmos
    Institute for Psychology, Hungarian Academy of Sciences, Budapest, Hungary
  • J.C. A. Read
    National Eye Institute, NIH, Bethesda, MD
  • Footnotes
    Commercial Relationships  L.R. Kozak, None; M. Castelo–Branco, None; G. Karmos, None; J.C.A. Read, None.
  • Footnotes
    Support  FCT SFRH/BD/13344/2003
Investigative Ophthalmology & Visual Science May 2004, Vol.45, 4366. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      L.R. Kozak, M. Castelo–Branco, G. Karmos, J.C. A. Read; A computational model of the motion aftereffect . Invest. Ophthalmol. Vis. Sci. 2004;45(13):4366.

      Download citation file:

      © ARVO (1962-2015); The Authors (2016-present)

  • Supplements

Abstract: : Introduction:Detailed information is available on the anatomy and circuitry of cortical visual areas underlying motion perception. Little is known however on how activity patterns are related to illusory motion perception. The motion aftereffect (MAE) is an important tool for investigating mechanisms involved in neuronal adaptation. However, most current models of the MAE include only a subset of the known physiology, and do not reproduce the phenomena observed with different stimuli, such as gratings, plaids or random–dot patterns. Purpose:We aimed to develop a biologically plausible model of the motion aftereffect, which is compatible with previous psychophysics, electrophysiology and imaging results. The model is aimed at describing both spatial and temporal activation patterns of the involved cortical areas. Methods:The model is developed as a set of differential equations representing cell activations and adaptation in visual areas V1 and MT (Grunewald 1995 NIPS 837–843). Our proposed model uses known biological parameters such as temporal and spatial frequency tuning (Simoncelli & Heeger Nat Neurosci 4:461–462) of visual motion selective neurons. The known connectivity between neurons in V1 and MT was also taken into account in our scheme. Model development and testing was done using Matlab®. Results obtained with the model were compared with biological data from our laboratory and the literature. Test stimuli included moving gratings, plaids and random dot patterns. Results:The model is able to simulate in a realistic manner both real and illusory motion related activation in the modeled brain areas (according to criteria of directionality and timing). The model was also able to generate activation patterns suggestive of the existence of component and pattern motion representations within MT. The results also support a possible neural implementation of the intersection of constraints rule (Adelson, Movshon Nature 300(5892):523–5.) Conclusions:Perception of real and illusory motion can be modeled simulating V1 and MT visual areas using cell properties and connectivity patterns based on biological data. Future work will attempt to simulate dynamical changes of speed perception of illusory motion.

Keywords: computational modeling • adaptation: motion • visual cortex 

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.