Jove
Visualize
Contact Us

Related Experiment Videos

Mixture models and the probabilistic structure of depth cues.

David C Knill1

  • 1Center for Visual Sciences, University of Rochester, 274 Meliora Hall, Rochester, NY 14627, USA. knill@cvs.rochester.edu

Vision Research
|March 18, 2003
PubMed
Summary

This study reveals how the brain uses prior knowledge, like texture properties, to interpret 3D visual scenes. It shows that visual perception flexibly adjusts constraints based on available image data for accurate depth perception.

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
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

Area of Science:

  • Visual perception
  • Computational neuroscience
  • Bayesian inference

Background:

  • Monocular depth cues rely on scene structure priors.
  • These priors are often mixture models for different scene categories.
  • Model selection is key to understanding cue integration.

Purpose of the Study:

  • To perform a Bayesian analysis of model selection for depth cues.
  • To investigate non-linear interactions between cues.
  • To derive and test psychophysical predictions for texture-based surface orientation perception.

Main Methods:

  • Bayesian analysis of mixture model priors.
  • Mathematical modeling of cue interactions.
  • Psychophysical experiments on texture-based surface orientation.

Related Experiment Videos

Main Results:

  • Mixed priors enable non-linear, cooperative cue interactions.
  • Single cues can determine appropriate scene constraints.
  • Visual system shows bias towards isotropic texture interpretation.
  • Isotropy constraint is turned off with sufficient data, favoring homogeneity.

Conclusions:

  • Human visual system acts like an optimal estimator with mixed priors on surface textures.
  • Perception adapts by adjusting constraints (e.g., isotropy) based on data.
  • This framework explains flexible and accurate depth perception from monocular cues.