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 stereo disparity using temporal smoothness constraints: a computational model

J V Stone1

  • 1Department of Psychology, University of Sheffield, Western Bank, England. stone@aivru.sheffield.ac.uk

Spatial Vision
|January 1, 1996
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

Where is the light? Bayesian perceptual priors for lighting direction.

Proceedings. Biological sciences·2009
Same author

Free-lunch learning: modeling spontaneous recovery of memory.

Neural computation·2006
Same author

Derivation of a clinical decision rule to guide the interhospital transfer of patients with blunt traumatic brain injury.

Emergency medicine journal : EMJ·2005
Same author

When is now? Perception of simultaneity.

Proceedings. Biological sciences·2002
Same author

Spatiotemporal independent component analysis of event-related fMRI data using skewed probability density functions.

NeuroImage·2002
Same author

Predicting spontaneous recovery of memory.

Nature·2001
Same journal

Comment on 'angle illusion on a picture's surface' by Hammad et al. (2008).

Spatial vision·2009
Same journal

Feature-based attentional modulation increases with stimulus separation in divided-attention tasks.

Spatial vision·2009
Same journal

Spatial distance between target and irrelevant patch modulates detection in a texture segmentation task.

Spatial vision·2009
Same journal

Inhibition related impairments of coherent motion perception in the attention-induced motion blindness paradigm.

Spatial vision·2009
Same journal

Recognition units in reading: backward masking experiments.

Spatial vision·2009
Same journal

Spatial-temporal modeling of interactive image interpretation.

Spatial vision·2009
See all related articles

This study introduces an unsupervised learning algorithm for stereo depth perception. The novel approach effectively estimates sub-pixel stereo disparity by assuming smooth depth variations over time, demonstrating strong generalization capabilities.

Area of Science:

  • Computer Vision
  • Machine Learning
  • Computational Neuroscience

Background:

  • Stereo vision is crucial for depth perception.
  • Unsupervised learning methods offer a data-driven approach to complex perceptual tasks.
  • Accurate estimation of stereo disparity is fundamental for 3D scene understanding.

Purpose of the Study:

  • To develop an unsupervised learning algorithm for estimating stereo disparity.
  • To leverage the assumption of smooth temporal depth variations for improved learning.
  • To evaluate the algorithm's performance on a hyperacuity task.

Main Methods:

  • An unsupervised learning algorithm was designed based on maximizing long-term output variance and minimizing short-term variance.
  • The learning rule combines anti-Hebbian and Hebbian weight changes across different timescales.

Related Experiment Videos

  • The algorithm was tested on estimating sub-pixel stereo disparity from temporal sequences of stereograms.

    • Main Results:

    • The algorithm successfully learned to estimate stereo disparity without labeled data.
    • Performance was demonstrated on a hyperacuity task, achieving precise sub-pixel disparity estimation.
    • The model exhibited generalization to novel image sequences without further training.

    Conclusions:

    • Unsupervised learning can effectively model stereo disparity based on temporal smoothness assumptions.
    • The proposed learning rule provides a biologically plausible mechanism for visual learning.
    • The algorithm shows promise for real-world applications requiring accurate 3D perception.