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 perceptually salient visual parameters using spatiotemporal smoothness constraints

J V Stone1

  • 1School of Cognitive and Computing Sciences, University of Sussex, UK.

Neural Computation
|October 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

A Model-Free Reinforcement Learning Implementation of Decision Making Under Uncertainty by Sequential Sampling.

Neural computation·2026
Same journal

DROP: Distributional and Regular Optimism and Pessimism for Reinforcement Learning.

Neural computation·2026
Same journal

Hierarchical Active Inference Using Successor Representations.

Neural computation·2026
Same journal

W-Kernel and Its Principal Space for Frequentist Evaluation of Bayesian Estimators.

Neural computation·2026
Same journal

A Hidden Markov Model-Inspired Sequence Classification Method for Hyperdimensional Computing.

Neural computation·2026
Same journal

Sparse Graphical Modeling for Electrophysiological Phase-Based Connectivity Using Circular Statistics.

Neural computation·2026
See all related articles

This study introduces an unsupervised learning model for vision tasks like surface depth extraction. It uses a novel learning rule to maximize long-term variance and minimize short-term variance, enabling effective stereo disparity learning.

Area of Science:

  • Computer Vision
  • Machine Learning
  • Computational Neuroscience

Background:

  • Unsupervised learning models are crucial for understanding low-level vision tasks.
  • Extracting surface depth and stereo disparity are fundamental challenges in computer vision.
  • Temporal smoothness is a key assumption in many biological and artificial perceptual systems.

Purpose of the Study:

  • To present a novel unsupervised learning model for low-level vision tasks, specifically surface depth extraction.
  • To derive a learning rule based on maximizing long-term and minimizing short-term variance of unit outputs.
  • To demonstrate the model's ability to learn stereo disparity from temporal sequences.

Main Methods:

  • Developed an unsupervised learning algorithm based on maximizing long-term and minimizing short-term variance.

Related Experiment Videos

  • The learning rule combines anti-Hebbian and Hebbian weight changes over different time scales.
  • Tested the model on temporal sequences of random-dot and gray-level stereograms with subpixel disparities.

    • Main Results:

    • The model successfully learned stereo disparity from temporal sequences without explicit supervision.
    • The algorithm demonstrated robustness to temporal discontinuities in disparity.
    • Generalization to unseen image sequences was achieved, indicating effective learning.

    Conclusions:

    • The proposed unsupervised learning approach is effective for low-level vision tasks like stereo disparity extraction.
    • The derived learning rule, balancing short-term and long-term variances, is a promising method for perceptual learning.
    • This work has implications for developing more sophisticated and adaptable artificial perceptual systems.