October 2020
Volume 20, Issue 11
Open Access
Vision Sciences Society Annual Meeting Abstract  |   October 2020
Visual predictions from physical relations
Author Affiliations
  • Alon Hafri
    Johns Hopkins University
  • Michael Bonner
    Johns Hopkins University
  • Chaz Firestone
    Johns Hopkins University
Journal of Vision October 2020, Vol.20, 1615. doi:https://doi.org/10.1167/jov.20.11.1615
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Alon Hafri, Michael Bonner, Chaz Firestone; Visual predictions from physical relations. Journal of Vision 2020;20(11):1615. doi: https://doi.org/10.1167/jov.20.11.1615.

      Download citation file:

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

  • Supplements

Understanding the world around us involves understanding not only which objects are where, but also how they relate—as when we see that one object is on, above, behind, or inside another. Some of these relations are physical, and play a special role in predicting the future state of a scene: If something is inside a cup—rather than occluded by it—we can assume that it will move if the cup moves. But beyond our ability to reason about what different physical relations entail, might they influence visual attention itself? Here, we ask whether the perception of physical relations automatically guides active maintenance of object identities—a core visual process by which the mind computes correspondence between current and previously seen objects. We turned the classic “object reviewing” paradigm into a “cups-and-balls” game in which participants rapidly responded to a target letter. On each trial, two balls with letters dropped into or behind two cups; then, the cups swapped places and disappeared, revealing the balls with letters. Crucially, the balls were equally likely to appear in the same or swapped locations, regardless of the relation (Containment vs. Occlusion). Nevertheless, we observed a “relation congruency” effect: For Containment more than for Occlusion, participants were faster to respond to the letters when they swapped positions along with the cups than when they remained in place. This effect held throughout the experimental session, suggesting that participants could not resist tracking the relations and their implied physical contingencies, even though these factors could not be used to predict final target locations. We propose that the mind updates object identities and their locations not only according to how individual objects move, but also according to the physical relations between them. In other words, relations—and their entailments about where things should be—influence core processes of attention and visual cognition.


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.