Perception organizes spatially distinct portions of the visual field into groups according to a set of Gestalt grouping cues including connection, common region, proximity, and similarity. Despite the long history of perceptual organization research, the mechanism underlying similarity grouping is rather poorly understood. Some cues, such as grouping by similar color, are so strong that observers may have the impression that the percept of multiple groups occurs spontaneously and in parallel. However, counter to this intuition, some research has suggested that similarity grouping is in fact feature selection – to group all the red objects together one must selectively attend to the color red. This proposal makes the striking prediction that to see multiple feature groups one has to attend to each feature serially, such that one cannot perceive a red group and a green group in the same instant. Here we provide additional evidence for this account from several different experimental paradigms including visual search and rapid enumeration. Our results suggest that simultaneous grouping by similarity may be merely an illusion -- rapid shifts of attention may create new feature groups on demand, leading to the impression that all groups are available simultaneously. 

Perceptual grouping is also an important factor in peripheral vision. The presence of grouping cues in peripheral vision strongly affects what one can perceive. Such grouping effects may allow us to distinguish between different models of peripheral vision, e.g. peripheral vision utilizes a predominantly feed-forward encoding of the visual input, or whether its mechanisms adapt to the stimulus organization. Understanding perceptual organization not only elucidates mechanisms of peripheral vision and attention, but is also important for many real-world applications, such as design of better information visualizations that allow the user to capture more information at a glance.

UPCOMING COG LUNCH TALKS:

11/28/17 - NO COG LUNCH

12/05/17 - Melissa Kline, Ph.D.

12/12/17 - Kara Weisman, Markman and Dweck Labs at Stanford University