Purchase this article with an account.
Orit Elgavi-Hershler, Shaul Hochstein; Vision at a glance: Faces do pop-out from a variety of other objects. Journal of Vision 2002;2(7):746. doi: https://doi.org/10.1167/2.7.746.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
It has been a basic tenet of the search literature that only elements that differ in a basic feature such as orientation or color would pop out from an array of distractors. More recent research has shown, however, that high-level concepts may pop out as well. Furthermore, experiments generally used a uniform array of distractors for search tests. In Experiment I subjects were presented with line drawings of faces, cars and houses as targets or distractors. Reaction time was independent of set size from 16 to 64 items/array when the target was a face and distractors were cars or houses, but increased with size when the target was a car or a house on a background of houses or cars, respectively. These results were obtained even when the distracting houses or cars differed in size and shape. We were concerned that the different outlines of round faces vs. linear houses and cars might serve as a low-level distinguishing feature. Experiment II therefore used a variety of real-life photographs containing faces and many other objects. Search for faces still had no set-size effect. The data indicate that important object categories are not detected by relatively slow, focused conjunction search, but by a faster system that is comparable to feature search. The results are in accord with recent RSVP experiments indicating that perception and categorization of high-level objects can be extremely rapid. Specifically, the results indicate that the visual system may be using a fast system with spread attention to find the gist of the scene, including object categories, as suggested by Reverse Hierarchy Theory (see Hochstein & Ahissar, VSS 2000).
This PDF is available to Subscribers Only