September 2021
Volume 21, Issue 9
Open Access
Vision Sciences Society Annual Meeting Abstract  |   September 2021
Physics 101: The visual systems ability to learn and integrate Newtonian predictions
Author Affiliations
  • Abdul Deeb
    Brown University
  • Fulvio Domini
    Brown University
Journal of Vision September 2021, Vol.21, 2620. doi:https://doi.org/10.1167/jov.21.9.2620
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Abdul Deeb, Fulvio Domini; Physics 101: The visual systems ability to learn and integrate Newtonian predictions. Journal of Vision 2021;21(9):2620. https://doi.org/10.1167/jov.21.9.2620.

      Download citation file:


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

      ×
  • Supplements
Abstract

The motions of objects in the environment reflect underlying dynamical constraints and regularities. Shepard (1984) first proposed that the perceptual system may have internalized these regularities to constrain how the system handles missing visual information. A recent body of work has demonstrated that not only is vision constrained by physical laws, but that these regularities can directly modulate motion perception when displayed events are inconsistent with events in the physical world (Deeb, Cesanek, Domini, 2021). In this current study we show that the visual system is capable of learning new physical regularities and can rationally incorporate this information with sensory data. We demonstrate that through multi-day exposure to non-Newtonian billiard ball collisions, the visual system can gradually relearn the internalized regularities used to modulate motion perception. Subjects were trained in a realistic 3D virtual environment and tasked with intercepting a launched ball. The launched ball moved along various deflection angles which all deviated from physics in a consistent, law-like manner. To assess subjects internalized visual predictions of object dynamics, we tested their ability to make absolute judgments pertaining to the launched ball’s deflection angle before and after training. Prior to non-Newtonian training, subject’s perceptual responses were similar to those found in Deeb et al. (2021), wherein subject’s responses reflected a rational inference combining a sensory estimate based on retinal image data with a Newtonian prediction based on pre-collision information. After the two consecutive days of training with new physical regularities we found a significant shift in participant’s responses toward the trained deflection angles. This however was not found in our control group who were trained with Newtonian stimuli. Thus, our results indicate that adaptive physical mappings, or priors, can be involved in motion perception and these mappings are not crystallized, but can be altered by experience.

×
×

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.

×