Abstract
Simons et al. (2006) showed that some visual transients can induce massive scene fading, in which a low pass filtered photographs of a natural scene fades to uniform luminance and color. At last year’s VSS, Francis & Kim (2010) explained these phenomena with simulations of the FAÇADE model (Grossberg & Mingolla, 1985a,b; Grossberg, 1997). In the model, oriented boundary responses to luminance edges undergo adaptation, and when the adaptation is substantial enough the weakened boundaries are no longer able to trap signals that represent brightness and color in a diffusive surface filling-in process. The effect of the boundary adaptation was frequently heightened by changes in the luminance edges, such as the offset of black disks and a global luminance decrement. Overall, the model’s behavior closely matched the findings reported by Simons et al. New simulations of the model explore the time course of the hypothesized adaptation and predict that the degree of fading should be positively related to the on time of the inducing black dots. We tested this prediction with a variation of the Simons et al. (2006) experiment, where observers made a judgment of surface homogeneity within a demarcated region after the offset of the dots. This new methodology avoids some possible hysteresis and criterion biases that might have been part of the Simons et al. study. As predicted by the model, reports of homogeneity increased with the duration of the black dots. The findings confirm the model’s prediction and lend validity to its explanation of the earlier data.
Meeting abstract presented at VSS 2012