Abstract
Last year Wede & Francis (VSS 2006) showed that Grossberg's FACADE theory explained why attentional focus to an inducing stimulus reduced the strength of the resulting negative afterimage. The model also correctly predicted that attentional focus would increase afterimage strength for a two-stimulus afterimage. In the model, attention is hypothesized to strengthen oriented responses and the resulting orthogonal after-responses that feed in to a filling-in stage. Orthogonal after-responses are unable to contain filling-in and so must fade away before a filled-in afterimage percept can be generated. We now show that the same mechanisms explain how the strength of afterimages vary with the spatial properties of the inducing stimulus. Georgeson & Turner (1985) reported that negative afterimages induced by square wave gratings were weaker than afterimages induced by sine wave gratings. We treated the model as a subject in the experiment and identified a cancellation stimulus that was able to minimize any resulting afterimage. The contrast of the cancellation image in the simulation highly correlated with contrast of the cancellation image from the experimental results, with sine wave inducers requiring a higher contrast cancellation stimulus. The model explains the data as due to differences in the strength of oriented responses. The sharp edges of a bar grating lead to stronger orientation after-responses, which lead to weaker negative afterimages and a weaker cancellation stimulus.