The contribution of borders to brightness and color perception

The light coming from a particular point in space does not determine the brightness and color appearance at that point. Light surrounding that point is also used to determine appearance. The influence of borders on the brightness and color appearance of neighboring regions is a fundamental example of how neural processes take into account surrounding light when determining perception of brightness and color. The first study in this project investigated the magnitude of the contribution of neural responses at the border to the color appearance of a uniform region, as well as the spatial integration properties of these neural responses. The second and third studies determined the neural locus underlying the (1) influence of borders on brightness perception, and (2) lateral interactions at borders mediating perception of temporally-varying brightness.;The first study examined how neural mechanisms responsive to either (i) borders and/or (ii) uniform light within borders contribute to color appearance. Specifically, this study determined what proportion of the color appearance of a uniform region could be explained solely by a neural response to its border. The border response accounted for 7% to 72% of the color appearance of a uniform region depending on the observer and condition, illustrating a critical importance of border responses for color perception. Furthermore, the border response did not simply reflect border contrast, as previous studies have suggested (Brenner et al., 2007). Instead, the results were consistent with a chromatic receptive field with band-pass spatial frequency tuning that integrates light at and around the border of chromatic regions.;The second study determined the neural locus underlying perception of the Craik-O'Brien-Cornsweet (COC) illusion, in which two adjacent equiluminant regions differ in brightness because of a light-dark border between them. Though perhaps the most well-known and quintessential demonstration of the importance of borders in brightness perception, the neural locus of the COC illusion is controversial. The neural basis of this illusion has been posited to reside in (i) monocular signals in the LGN (Anderson et al., 2009), (ii) lower-lever visual cortical areas, i.e. V1 and V2 (Boyaci et al., 2007; Roe et al., 2005), or (iii) higher-level visual cortical areas (Perna et al., 2003). In three experiments here, the neural mechanism underlying perception of the COC illusion was localized to a binocular neural site. No evidence was found in support of a monocular/LGN locus. A monocular COC border was neither necessary nor sufficient to produce the COC illusion. The results imply that the illusion depends on a neural representation at a cortical locus after both binocular combination and binocular rivalry.;The third study determined the locus of neural mechanisms mediating lateral interactions in perception of temporally-varying brightness. Perception of a temporally-varying light is strongly affected by temporal variation within a surrounding field. The neural mechanism mediating this perceived lateral interaction has been posited to be center-surround antagonism in the LGN, i.e. monocular signals (Kremers et al., 2004) or, alternatively, at a cortical (binocular) level (D'Antona & Shevell, 2007). To determine the neural locus, the contribution of monocular and/or binocular components to these lateral interactions were either measured directly or estimated from a simple model. The results revealed both a weak monocular (plausibly LGN) and large binocular (central) component of lateral interaction. The monocular component was relatively constant as a function of temporal frequency (from 3.125 to 12.5 Hz), while the binocular component showed low-pass temporal-frequency tuning. These findings are consistent with two separate neural sites (monocular and binocular) underlying perceived temporal variation in context, with each site having a distinct strength and temporal-frequency tuning.