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

Uncertainty in visual processes predicts geometrical optical illusions.

Cornelia Fermüller1, Henrik Malm

  • 1Department of Computer Science, Computer Vision Laboratory, Center for Automation Research, Institute for Advanced Computer Studies, University of Maryland, College Park, MD 20742-3275, USA. fer@cfar.umd.edu

Vision Research
|January 31, 2004
PubMed
Summary

Optical illusions arise from statistical biases in visual computations, where image feature estimation is affected by noise. This leads to misperceptions, demonstrating a general uncertainty principle in vision systems.

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

Single-Step Latent Diffusion for Underwater Image Restoration.

IEEE transactions on pattern analysis and machine intelligence·2025
Same author

Minimal perception: enabling autonomy in resource-constrained robots.

Frontiers in robotics and AI·2024
Same author

Exploring synthetic datasets for computer-aided detection: a case study using phantom scan data for enhanced lung nodule false positive reduction.

Journal of medical imaging (Bellingham, Wash.)·2024
Same author

Microsaccade-inspired event camera for robotics.

Science robotics·2024
Same author

Ajna: Generalized deep uncertainty for minimal perception on parsimonious robots.

Science robotics·2023
Same author

Semi-supervised training using cooperative labeling of weakly annotated data for nodule detection in chest CT.

Medical physics·2023

Area of Science:

  • Visual perception
  • Computational neuroscience
  • Image processing

Background:

  • Visual interpretation relies on deriving image features like lines and motion.
  • Image formation and processing are subject to noise and uncertainty, leading to estimation biases.
  • These biases can cause erroneous perception of feature locations and altered pattern appearance.

Purpose of the Study:

  • To propose that optical illusions stem from the statistics of visual computations.
  • To explain how noise and uncertainty in image feature estimation lead to perceptual biases.
  • To introduce a general uncertainty principle governing vision systems.

Main Methods:

  • Analysis of visual computations and feature estimation processes.
  • Examination of noise and bias in image processing.

Related Experiment Videos

  • Theoretical framework development based on statistical properties of vision.

    • Main Results:

    • Geometrical optical illusions and motion-based illusory patterns are linked to visual computation statistics.
    • Bias in feature estimation, though often small, is highly pronounced in illusory patterns.
    • Optical illusions are presented as artifacts of an inherent uncertainty principle in vision.

    Conclusions:

    • A general uncertainty principle governs vision systems.
    • Optical illusions are a consequence of biases in visual feature estimation due to noise.
    • Understanding these statistical properties offers insights into visual perception mechanisms.