Abstract
At last year’s VSS, Francis & Kim (2010) demonstrated that simulations of a cortical model of visual perception explained the large scale induced fading effects reported by Simons et al. (2006). The main model feature was that oriented boundary responses to luminance edges were subject to adaptation, and when the adaptation was substantial enough the weakened boundaries allowed color and brightness signals to spread across surfaces in a diffusive filling-in process. Changes in the image (such as the disappearance of high contrast dots or a global decrease in luminance) exacerbated these effects by introducing an orientation after-response among boundary representations. In new model simulations we explored this explanation by varying the duration of high contrast dots and measuring the amount of model-predicted fading. As the duration of the dots increased over several seconds, the model predicts stronger adaptation and more induced fading at dot offset. These predictions are in contrast to a possible alternative hypothesis that the transient signals at dot offset are critical for induced fading. Given the durations of the dots, this alternative explanation would not predict any difference because the transient offset responses should be equivalent. In further explorations of the model, we noted that the model suggests that there should be substantial variation in fading across the images used by Simons et al. (2006). Overall, the model both accounts for previous findings and suggests novel properties of induced fading that can be tested in future experiments. Because the model has previously been used to explain a wide variety of related phenomena (including afterimages and visual persistence), it seems that a relatively small set of mechanisms in visual cortex provide a unified explanation of disparate phenomena.
Meeting abstract presented at VSS 2012