Guided search: An alternative to the feature integration model for visual search.

Abstract

Subjects searched sets of items for targets defined by conjunctions of color and form, color and orientation, or color and size. Set size was varied and reaction times (RT) were measured. For many unpracticed subjects, the slopes of the resulting RT x Set Size functions are too shallow to be consistent with Treisman's feature integration model, which proposes serial, self-terminating search for conjunctions. Searches for triple conjunctions (Color x Size x Form) are easier than searches for standard conjunctions and can be independent of set size. A guided search model similar to Hoffman's (1979) two-stage model can account for these data. In the model, parallel processes use information about simple features to guide attention in the search for conjunctions. Triple conjunctions are found more efficiently than standard conjunctions because three parallel processes can guide attention more effectively than two.

Task-Dependent Learning of Attention

“…However, these studies claim different human performances in conjunction search. The experiments done by Wolfe et al (1989) and Treisman and Sato (1990) suggest that the standard conjunction target is detected in parallel. Quinlan and Humphreys (1987), on the other hand, state that subjects perform a serial search to detect the target.…”

“…, the time needed to move attention from the current location to the new position. Based on values extracted from psychological experiments (Treisman and Sato, 1990;Wolfe et al, 1989), the initial response time has been set to 300 ms and the duration of a shift of attention has been set to 60 ms. With these definitions, the model's reaction time is defined as…”

“…A visual search paradigm that is often used (Quinlan and Humphreys, 1987;Treisman and Sato, 1990;Wolfe et al, 1989) is feature and conjunction search. In a feature search task the target differs in one and the same feature from all distractors.…”

“…All studies on feature and conjunction search (Quinlan and Humphreys, 1987;Treisman and Sato, 1990;Wolfe et al , 1989) report that in feature search the target is detected in parallel. However, these studies claim different human performances in conjunction search.…”

“…Because humans do not have maps in the preattentive stage for combinations of features of the color, orientation, and size dimensions (Quinlan and Humphreys, 1987;Treisman and Sato, 1990;Wolfe et al, 1989) , the target and distractors must be separated purely based on features, not on conjunctions of features . Therefore, humans use top-down information to select features that separates the target from (some or all) distractors, while ignoring features that are shared by both the target and all distractors.…”