ABSTRACT
The visual brain is tasked with segmenting visual input into a structured scene of coherent objects. Figures that produce the perception of illusory contours reveal minimal conditions necessary for the visual system to perform figure-ground segmentation, but it is unclear what role visual attention plays in such perceptual organisation. Here we tested whether illusory contours are formed under the guidance of voluntary attention by exploiting a novel variant of the classical Kanizsa figure. We used a psychophysical response classification technique to quantify spatial structures guiding perceptual decisions, and found that illusory contour formation was contingent on attended elements of the figure. However, the strength of these illusory contours was constrained by form implied by unattended figural elements. Our data thus reveal an interplay of top-down and bottom-up processes in figure-ground segmentation: while attention is not necessary for illusory contour formation, under some conditions it is sufficient.