Abstract
Visual search performance is aided by knowledge of scene context and likely target positioning (Castelhano & Henderson, 2007; Neider & Zelinsky, 2006). In a recent study, Pereira and Castelhano (2017) demonstrated that attention is differentially deployed depending on the target-relevance of each scene region: in an abrupt-onset paradigm, target-relevant distractors were fixated upon and saccaded towards significantly more often than target-irrelevant distractors. In the present study, we examined whether the visual features of distractors influence attentional deployment over and above the spatial relevance of their positions. Distractors were placed in scene regions that were operationalized as either target-relevant or target-irrelevant, and were either visually similar or dissimilar to the target object. Participants saccaded towards and fixated upon target-relevant distractors significantly more often than target-irrelevant distractors. Interestingly, visual target-distractor similarity did not have an effect: only distractors appearing within target-relevant regions reliably attracted attention, regardless of their visual similarity to the target. These findings suggest that attention during search is distributed based on likely target positioning and, surprisingly, that attentional capture within scenes may be better predicted by spatial relevance than by visual feature similarity. However, previous research has also demonstrated that distractors are more likely to capture attention when they share categorical features with the target (Wyble, Folk, & Potter, 2013). We will further examine the potential interaction between the spatial relevance of distractor positioning and categorical target-distractor similarity, in order to assess the extent to which spatial relevance and distractor features differentially predict attentional deployment in search through real-world scenes.
Meeting abstract presented at VSS 2018