Explaining the effects of distractor statistics in visual search

Abstract

Visual search, the task of detecting or locating target items amongst distractor items in a visual scene, is an important function for animals and humans. Different theoretical accounts make differing predictions for the effects of distractor statistics. Here we use a task in which we parametrically vary distractor items, allowing for a simultaneously fine-grained and comprehensive study of distractor statistics. We found effects of target-distractor similarity, distractor variability, and an interaction between the two, although the effect of the interaction on performance differed from the one expected. To explain these findings, we constructed computational process models that make trial-by-trial predictions for behaviour based on the full set of stimuli in a trial. These models, including a Bayesian observer model, provided excellent accounts of both the qualitative and quantitative effects of distractor statistics, as well as of the effect of changing the statistics of the environment (in the form of distractors being drawn from a different distribution). We conclude with a broader discussion of the role of computational process models in the understanding of visual search. claim that Treisman and Gelade (1980) only explored part of the stimulus space spanned by targetdistractor similarity and distractor variability; the appearance of a dichotomy would then stem from the use of stimuli drawn from two distinct clusters in this space.While Duncan and Humphreys (1989) made a valuable contribution in suggesting that performance likely lies on a continuum, their exploration of this claim was necessarily limited by the stimuli that they used. Using letters and joined lines, they could not parametrically vary properties of these stimuli along an easily quantified dimension. Therefore, it could be that Duncan and Humphreys (1989), while describing performance in a larger area of stimulus space than Treisman and Gelade (1980), missed areas of the space, along with distinctive qualitative effects.Other visual search researchers have used stimuli that can easily be parametrically varied (Palmer,

The psychophysics of visual search with heterogeneous distractors

“…As we said before, the optimal-observer model only differed from the localization one in the decision rule stage. We were also able to capture the detection data relatively well with the optimal-observer model (Figure 6), in line with Mazyar et al (2012) and Calder-Travis and Ma (2020). These fits are satisfactory; the data was captured even better by including a decision noise parameter (Figure 7).…”

“…We defined a rich set of five summary statistics (Table 2). Summary statistics A - D overlap with the ones from Calder-Travis and Ma (2020). We characterized performance in localization according to all five summary statistics, and in detection according to A-D, as E was only applicable to localization.…”

“…Distributions of targets and distractors other than the uniform would yield different patterns of these correlations. To illustrate this, Table A1B shows the case when target and distractors follow a von Mises distribution with mean 0 and concentration parameter 1.5, as in Calder-Travis and Ma (2020). The Spearman correlations between min T-D difference and T-D mean are positive and only mildly decrease with increasing set sizes, while the min T-D difference and distractor variance are not significantly correlated.…”

“…This pattern of results held true regardless of whether the identity of the target was revealed before or after the search array, meaning for both visual search in perception and in memory. Calder-Travis and Ma (2020) extended the study of target detection in visual search with heterogeneous distractors by employing two types of distractor distributions; they also found that an optimal-observer model captured the data well.…”

“…Therefore, we characterized search behavior as a function of the following summary statistics which naturally emerge from the uniform distribution, computed on a trial-by-trial basis: the difference between the target and the most similar distractor (Mazyar et al, 2012, 2013) the difference between the target and the mean of the distractors, and distractor variance. Only one study so far has investigated all these summary statistics (Calder-Travis & Ma, 2020). Because we use a unidimensional, continuous stimulus variable (orientation), these quantities can be defined exactly and unambiguously.…”