Purchase this article with an account.
Anton Gura, Igor Utochkin; The perception of variety in color segmented sets. Journal of Vision 2014;14(10):877. https://doi.org/10.1167/14.10.877.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Observers can extract summary variety statistics from large sets of objects (Utochkin & Gura, VSS, 2014, this issue). Here we tried to find out how the visual system analyses visual variety statistics of segmented groups of objects. In Experiment 1 we tested whether variety statistics can be extracted from spatially overlapping sets segmented by color, as they are extracted from spatially separated sets (see Utochkin & Gura, this issue, Experiment 1). Observers were presented with sets of red and blue differently sized circles and had to determine which of sets had been more various in terms of size. The sets could be identical, or different by bandwidth, heterogeneity, or both. The results showed that variety discrimination was as efficient as in spatially separated sets indicating that the visual system is able to represent variety independently for different features. In Experiment 2, we tested whether segmentation of a set into heterogeneous subsets affects overall variety representation of that set. Observers were presented with sets of differently sized circles to the right and left from fixation. One of those sets was always colored homogeneously (in either red, green, or blue) and the other was always heterogeneous (including all three colors). Observers had to respond which of two sets was more various in terms of size, regardless of color variation. As in Experiment 1, size bandwidth and heterogeneity were manipulated. The results showed no effect of color heterogeneity on variety discrimination as compared to completely homogeneous sets (see Utochkin & Gura, this issue, Experiment 1). This indicates that variety information can be encoded along a sensory dimension (such as size) independently from variation of other features. This finding is in line with a standard two-stage model of vision considering ensemble summary statistics to be an output of early feature processing (Treisman, 2006).
Meeting abstract presented at VSS 2014
This PDF is available to Subscribers Only