Abstract
The visual system is remarkably efficient at extracting statistical ensembles from objects in the environment, such as the mean size or orientation. Yet at any given time, multiple groups of objects can be randomly distributed over space. Thus, the challenge for the visual system is to summarize over multiple intermixed sets at once. What is the limit of the ability to perceive multiple ensembles? In a series of experiments, participants viewed an array of 1 to 8 spatially intermixed sets of circles for 1000ms in each trial. Each set contained four circles in the same colors but with different sizes. A probed set was randomly chosen from the array, and was either pre-cued or post-cued. Participants estimated the mean size of the probed set. Fitting a uniform-normal mixture model to the error distribution for each number of set, we found a four-set limit of ensemble perception: observers could reliably estimate the mean size of circles from maximally four sets (Experiment 1). Importantly, their performance was unlikely to be driven by a subsampling strategy (Experiment 2). By extending exposure durations to 1500ms and 2000ms, we found that the estimation of mean size might be constrained by internal capacity constraints even when the constraint from processing speed was eliminated (Experiment 3). In addition, ensemble perception may be limited by the storage capacity of visual working memory (Experiment 4). Finally, the capacity for ensemble perception does not seem to be influenced by the imprecise representations of individual circles in the set (Experiment 5). Overall, our findings suggest that ensemble perception can operate over multiple intermixed sets, but with a four-set capacity limit. This result converges with previously observed limits in visual working memory, attention, and enumeration. The convergence implies that different forms of visual processes may share a common capacity constraint.
Meeting abstract presented at VSS 2016