Abstract
Hue ensemble percepts can be formed over a hue distribution that is spatially intermixed with a distractor hue distribution. We asked whether a color category boundary aids in segregating hue distributions from one another when forming ensemble percepts. Stimulus colors were selected from a hue circle in CIELUV color space. Color category boundaries were estimated with a categorization task. In the main task, test and comparison stimuli consisted of 18 circular patches (0.5° in diameter) randomly placed on a 6-by-6 grid (1° center-to-center spacing), with hues drawn from a uniform distribution with a 25° hue angle range. Observers (N=15) were split into two groups where test distribution mean hue angle was -/+ 22.5° from the green-blue boundary. There were three conditions: a baseline with no distractors, and two conditions with distractor patches filling the other 18 spaces in the test stimulus grid. The distractor distribution mean was 45° from test mean hue, either “across” the green-blue boundary or within the same category as the test. Comparison mean hue was varied around the test mean hue. The test and comparison were presented in two temporal intervals, and the observer responded whether the hue distribution in the second interval was “yellower/greener” or “bluer/purpler” than in the first interval, ignoring distractor patches. Psychometric functions were fit to proportion of ”bluer/purpler” responses to estimate the perceptual bias and discrimination threshold. In distractor conditions, perceived hue was biased towards distractor hues with an elevated discrimination threshold compared to baseline. On average, there was no difference in bias or threshold between “across” and “within” distractor conditions, suggesting that color category boundaries do not aid segregation of hue ensembles. Some observers, however, showed a considerably stronger bias in one, most often in the “across”, condition. Possible sources of these individual differences will be discussed.