Abstract
Relative phase of compound grating might be a useful cue to the detection of edges and lines, for which the visual system may have developed specialised detectors. Previous studies have aimed to characterise such detectors by measuring the sensitivity of human observers to changes in relative phase (e.g. Burr et al., 1989, Vis Res 29: 391-6; Huang et al., 2006, Vis Res 46: 2069-2081). However, measuring the sensitivity of the overall system does not, tell us about the selectivity of the underlying mechanisms; whether or not selective detectors exist. Here we used compound adaptation to study the detection of relative phase. We created stimuli by combining gratings of f, 3f, 5f… in phases between 0° (zero-crossing align; edge phase) and 90° (peaks/troughs align; line phase). Participants were adapted to edge- and line-phase stimuli simultaneously (one in each hemi-field). The absolute phase of the stimuli was jittered. We then measured whether this adaptation had any differential effect in the two hemi-fields, in terms of the perceived form and contrast of subsequently presented stimuli. Any such differences must result, not from adaptation to the component gratings, but to the compound pattern, since the component gratings in each hemi-field were identical. There was a shift in the point of subjective equality (PSE) for relative phase; intermediate-phase stimuli were perceived as more “edge-like” after adaptation to a line-phase stimulus and vice versa. Similarly, there were PSE shifts in the contrast domain, indicating a lower apparent contrast of probes with the same relative phase as the adaptor. These effects can only be attributed to the differences in the relative phase (alignment) of the component sinusoids and are consistent with the existence of a neural mechanism responding selectively to stimuli of particular relative phases.
work was supported by The Wellcome Trust (WT085444).