Purchase this article with an account.
Erika Wells, Andrew Leber, Erin Kuta; Attentional Set Produces an Inhibitory Surround in Color Space. Journal of Vision 2011;11(11):93. doi: https://doi.org/10.1167/11.11.93.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
How is an attentional set for color instantiated? By one account a particular region of color space is enhanced focally, with this enhancement decreasing as the distance in color space from the focal region increases. Alternatively, feature inhibition may be applied to intermediate distances surrounding the enhanced region similar to inhibitory surround effects in orientation feature space (Tombu & Tsotsos, 2008). We tested these two alternatives using an attention capture task in the color domain. Method: Participants searched for a color outline circle target (e.g., red) among 3 gray non_target circles and reported whether a gap appeared at the top or bottom of the target. The target color was fixed across trials for each participant (determined from a set of L*a*b color coordinates centered at a = 10, b = 10, r = 50). On each trial, a set of small, equally colored distractor dots briefly surrounded either the target or one of the distractor circles, prior to the target onset (Folk, Remington & Johnston, 1992). The distractor's distance in L*a*b color space from the target was parametrically varied between 0 and 180 degrees of polar angle, allowing us to quantify the extent of attention capture as a function of color distance. Results/Discussion: We found maximal interference by distractors sharing the same color as the target. Capture then decreased as the feature distance between the target and distractor increased. However, this decrease reached a minima at intermediate color distances (60 and 90 degrees), beyond which point interference increased. The pattern of interference was best categorized by a quadratic function. These results indicate that attentional set for color consists of both a central enhancement and an inhibitory surround, allowing efficient tuning to the behaviorally relevant feature.
This PDF is available to Subscribers Only