A signal detection model predicts the effects of set size on visual search accuracy for feature, conjunction, triple conjunction, and disjunction displays

Abstract

Recently,quantitative models based on signaldetection theory have been successfully applied to the prediction of human accuracy in visual search for a target that differs from distractors along a single attribute (feature search). The present paper extends these models for visual search accuracy to multidimensional search displays in which the target differs from the distractors along more than one feature dimension (conjunction, disjunction, and triple conjunction displays). The model assumes that each element in the display elicits a noisy representation for each of the relevant feature dimensions. The observer combines the representations across feature dimensions to obtain a single decision variable, and the stimulus with the maximum value determines the response. The model accurately predicts human experimental data on visual search accuracy in conjunctions and disjunctions of contrast and orientation. The model accounts for performance degradation without resorting to a limited-capacity spatially localized and temporally serial mechanism by which to bind information across feature dimensions.Visual search for a target among a set of distractors has been extensively studied by a large number of investigators, Typically, the observer's reaction time for finding the target is measured as a function of the number of distractors (set size) in the display. When the target and the distractors differ along one physical dimension or stimulus attribute (e.g., length, orientation, color, brightness, etc.), the search task is known as a feature search. A comSome of the results of this paper were presented at

Attention capacity and task difficulty in visual search

“…However, the present results (along with those of Eckstein et al, 2000;Geisler & Chou, 1995, and others) suggest that a rather different hierarchy of tasks may characterize the relative visual-attention demands of different kinds of visual search tasks.…”

“…Therefore, it appears that this kind of search is indeed fundamentally different from the other two kinds tested here (Wolfe, 1998b). Inefficient feature search arises when the system need to accumulate information for a longer period of time in order to distinguish signal from noise (Eckstein, Thomas, Palmer, & Shimozaki, 2000;Lu & Dosher, 1998). In spatial configuration search, however, the inefficiency is in the selection time.…”

Computational Models of Human Visual Attention and Their Implementations: A Survey

“…The major contribution of Eckstein et al [47] is that they were the first to clarify the computational mechanism of those relationships.…”

“…Here we assume that people may observe a different map from the saliency map because of the internal noise of the visual cortex, where every internal noise is emitted from an independent Gaussian distribution [47]. The observed saliency map is called a stochastic saliency map, and each of its pixel values is called a stochastic saliency.…”

“…Eckstein et al [47], [48] applied the signal detection theory to visual attention modeling, and described a mechanism that causes attention ambiguity using only bottomup processing. The idea behind the signal detection theory can be intuitively explained with the help of Fig.…”