September 2011
Volume 11, Issue 11
Vision Sciences Society Annual Meeting Abstract  |   September 2011
Automatic feature-based grouping during multiple object tracking
Author Affiliations
  • Everett Mettler
    University of California, Los Angeles
  • Brian Keane
    University of Dentistry and Medicine of New Jersey
    Rutgers University Center for Cognitive Science
  • Genna Erlikhman
    University of California, Los Angeles
  • Todd Horowitz
    Visual Attention Laboratory, Brigham and Women's Hospital
    Department of Opthamology, Harvard Medical School
  • Philip Kellman
    University of California, Los Angeles
Journal of Vision September 2011, Vol.11, 287. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Everett Mettler, Brian Keane, Genna Erlikhman, Todd Horowitz, Philip Kellman; Automatic feature-based grouping during multiple object tracking. Journal of Vision 2011;11(11):287.

      Download citation file:

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

  • Supplements

Purpose: Prior studies (Keane, Mettler, Tsoi, & Kellman, 2010) have shown that the features of targets and distractors can be utilized to form illusory contours during multiple object tracking. When contours connect targets, tracking improves; when contours connect targets and distractors, performance worsens. Can features besides those that elicit interpolation also lead to automatic grouping? Method: Displays contained four pairs of objects, one pair in each quadrant. Observers tracked one target in each pair. We examined grouping in the context of seven additional potential grouping factors: color, contrast polarity, orientation, size, shape, stereoscopic depth, and a combination (size+shape+color). Displays were created using two feature values of each grouping factor (e.g., red and green for color). All objects were identical at target designation and at response; distinctive features were only visible during the motion phase. There were two conditions in Experiment 1. For the target-group (TG) condition, all targets had one feature (e.g., red) and all distractors had the remaining feature (e.g., green). For the target-distractor group (TDG) condition, half the targets and half the distractors had one feature and the remaining objects had the other feature. In Experiment 2, we compared the TDG to a No-group baseline condition in which all objects were identical at each moment. Results: In Experiment 1, the TG condition was better than the TDG condition for all grouping factors (all ps < 0.001) except orientation. In Experiment 2, performance in the TDG condition was worse than the No-group condition for all grouping factors (all ps < 0.005) except for orientation, depth, and contrast polarity. Conclusion: Observers used grouping factors in both experiments even though they were not instructed to. In Experiment 2, they used grouping factors even though it hurt their performance. In addition to interpolation, color, size and shape can lead to automatic grouping during attentional tracking.


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.