Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Learning Bayesian priors for depth perception.

David C Knill1

  • 1Center for Visual Science, University of Rochester, Rochester, NY, USA. knill@cvs.rochester.edu

Journal of Vision
|August 10, 2007
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Recovering stereo vision by squashing virtual bugs in a virtual reality environment.

Philosophical transactions of the Royal Society of London. Series B, Biological sciences·2016
Same author

Learning and inference using complex generative models in a spatial localization task.

Journal of vision·2016
Same author

Unifying account of visual motion and position perception.

Proceedings of the National Academy of Sciences of the United States of America·2015
Same author

Stereopsis and amblyopia: A mini-review.

Vision research·2015
Same author

Kinesthesis can make an invisible hand visible.

Psychological science·2013
Same author

The brain uses adaptive internal models of scene statistics for sensorimotor estimation and planning.

Proceedings of the National Academy of Sciences of the United States of America·2013
Same journal

Analysis of human visual experience data.

Journal of vision·2026
Same journal

Pyramid-based Bayesian modeling for high-resolution behavioral analysis.

Journal of vision·2026
Same journal

Sensation without perception: The white whale effect and perceptual blindness in autonomous vehicles.

Journal of vision·2026
Same journal

Gaze behavior during closed-captioned movie viewing adapts to absent audio through more frequent switching between text and scene.

Journal of vision·2026
Same journal

In pursuit of saccade awareness: Limited volitional control and minimal conscious access to catch-up saccades during smooth pursuit eye movements.

Journal of vision·2026
Same journal

Dissociable effects of element-lifetime and stimulus-duration on local and global motion processing: An equivalent noise study.

Journal of vision·2026
See all related articles

The visual system adapts its depth perception by learning statistical regularities from visual input. This study shows that exposure to random ellipses reduces reliance on a prior bias for interpreting shapes as slanted circles.

Area of Science:

  • Perceptual Science
  • Computational Neuroscience
  • Computer Vision

Background:

  • Understanding how the visual system infers 3D depth from 2D images is a key challenge.
  • The visual system relies on statistical regularities, like symmetry, for interpreting depth cues.

Purpose of the Study:

  • To investigate if the visual system adapts its statistical model for estimating 3D surface orientation.
  • To determine if exposure to non-elliptical shapes alters the interpretation of depth cues.

Main Methods:

  • Subjects judged the slant of stereoscopically viewed ellipses under varying conditions.
  • Experiment 1 involved haptic feedback; Experiment 2 assessed learning without feedback.
  • A Bayesian model was used to analyze cue integration and prior adaptation.

Related Experiment Videos

Main Results:

  • When presented with random ellipses, subjects progressively reduced reliance on the 'slanted circle' prior for depth judgments.
  • This adaptation occurred even without explicit feedback on surface orientation.
  • The visual system adjusts its interpretation of pictorial depth cues based on environmental statistics.

Conclusions:

  • The visual system dynamically adapts its priors for depth perception based on statistical regularities in the visual environment.
  • Learning statistical regularities, even from non-canonical shapes, is crucial for accurate 3D surface orientation estimation.
  • This adaptive mechanism allows for flexible and robust depth perception across different visual scenes.