Abstract
The light field in a scene describes the light flow through each point in space from every direction. In previous work, we found that the visual system interpolates information about the light field across empty regions to an isolated test patch (Kim, Schüür & Maloney, 2013). Here, we examined whether we could disrupt the interpolation process by surrounding flankers with a uniform white border (by analogy with previous work in lightness perception; Gilchrist et al., 1999). Participants viewed a target (diamond) situated between two flankers (two large, white blobs) above and below the target for 750ms. The flankers were lit by a yellow proximal and a blue collimated light source. The task was to indicate whether the target needed more yellow or more blue light in order to appear neutral (white). The flankers appeared without borders in one condition (control) while, in two other conditions they were surrounded by borders. In one condition, the borders were lit as the flankers while in the other condition, the borders were not shaded. Neither the shaded nor the unshaded borders disrupted light field interpolation in four out of six observers. For the remaining two, we found failure of interpolation in all three conditions. A white border may not be sufficient to disrupt light field interpolation; light field interpolation is remarkably robust.
Meeting abstract presented at VSS 2014