September 2011
Volume 11, Issue 11
Vision Sciences Society Annual Meeting Abstract  |   September 2011
Feature selection as a mechanism for color grouping
Author Affiliations
  • Derek Tam
    Department of Psychology, Northwestern University, USA
  • Brian Levinthal
    Department of Psychology, Northwestern University, USA
  • Steven Franconeri
    Department of Psychology, Northwestern University, USA
Journal of Vision September 2011, Vol.11, 1082. doi:10.1167/11.11.1082
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Derek Tam, Brian Levinthal, Steven Franconeri; Feature selection as a mechanism for color grouping. Journal of Vision 2011;11(11):1082. doi: 10.1167/11.11.1082.

      Download citation file:

      © ARVO (1962-2015); The Authors (2016-present)

  • Supplements

The visual system forms groups out of objects that share similar features, such as color, shape, or orientation. One proposed mechanism for this similarity grouping is selection of a given feature, which should increase activation in visual field locations that contain that feature (e.g. red or vertical). Because these locations a selected together, they appear to belong together (e.g., Huang & Pashler, 2007). This mechanism makes a unique and surprising prediction - because groups are defined by selected regions, only one group can exist at a time. When we feel that we are grouping two independent sets simultaneously (e.g. two red objects and two green objects), this must be an illusion caused by rapid switching between features. We tested this idea with a visual search tasks for groups among other groups - if only one group can be created at a time, these searches should be extremely ine fficient. As a control to ensure that inefficient search was due to grouping processes per se, we also asked participants to perform similar tasks with connected groups, which are known to be constructed in parallel. In one experiment, participants were extremely slow to locate an unmatched pair of colored squares (e.g., a red and blue square) among matched pairs (e.g., two green squares) (77 ms/pair). In a control condition, the objects within each pair were moved closer together so that they touched and created a color contrast edge. Search was now highly efficient (16 ms/pair). In a second experiment, participants were asked to search for a vertical group of dots among horizontal groups. Search was again inefficient (27 ms/group) but was significantly faster when groups were connected by lines (15 ms/group). Both of these results are consistent with the possibility that our feeling of grouping objects with similar features may be due to selection of those features.

Northwestern University Summer and Academic-Year Undergraduate Research Grants. 

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.