Purchase this article with an account.
A.J Ayeni, William Harrison, Peter Bex; Filling-in of Kanizsa-style illusory figures is under top-down control. Journal of Vision 2015;15(12):532. doi: https://doi.org/10.1167/15.12.532.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Illusory Kanizsa figures demonstrate the visual system’s ability to interpolate shape and structure from fragmented information. For example, on a gray background, white “pacman” inducers arranged at the corners of an imaginary triangle produce an apparent occluding surface that is darker than the background, whereas black pacmen produce an apparently lighter occluding surface. Although such “filling-in” has been accounted for by the bottom-up properties of early visual cortex, we sought to investigate the extent to which top-down attention can influence the computations underlying visual interpolation. We thus developed a perceptually bi-stable hexagram figure composed of two inverted and spatially overlapping Kanizsa triangles on a gray background (20 cd/m2). The pacmen defining one triangle were white (40 cd/m2), and the others were black (0 cd/m2). An observer’s task was to report whether only one of the two illusory figures was lighter or darker than a luminance-defined matching triangle presented on a background of white noise, with match luminance adjusted according to an adaptive staircase. To draw observers’ attention to one of the two illusory figures, they were instructed to judge only the illusory triangle that matched the orientation of the luminance-defined triangle, which was chosen randomly trial-by-trial. We therefore directed observers to attend to an illusory triangle defined by white (or black) inducers endogenously, but not by referring to the color of the inducers. For 15 observers, perceptual reports strongly depended on the pacman color of the attended illusory triangle, despite both sets of pacmen being present for all trials. Because the illusory triangles spatially overlapped, these opposite changes in the perceived lightness of the surface could only be due to our attention manipulation. We thus show that, when at least two percepts are equally plausible, visual attention ultimately guides the perceptual experience of filling-in.
Meeting abstract presented at VSS 2015
This PDF is available to Subscribers Only