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

Modelling symmetry detection with back-propagation networks

C Latimer1, W Joung, C Stevens

  • 1Department of Psychology, University of Sydney, New South Wates, Australia.

Spatial Vision
|January 1, 1994
PubMed
Summary

This study found that both humans and artificial neural networks detect vertical and horizontal symmetries faster after specific training. This suggests early visual experience influences symmetry perception speed.

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

Association between ten complications and mortality : insights from the Dutch Hip Fracture Audit.

The bone & joint journal·2026
Same author

The global extent of the grassland biome and implications for the terrestrial carbon sink.

Nature ecology & evolution·2026
Same author

Mobility and living status at discharge and after three-months for extramedullary versus intramedullary fixation of AO type 31-A1 trochanteric fractures; an analysis of Dutch hip fracture audit data.

European journal of trauma and emergency surgery : official publication of the European Trauma Society·2025
Same author

Crevasse refreezing and signatures of retreat observed at Kamb Ice Stream grounding zone.

Nature geoscience·2023
Same author

SDF to the rescue.

British dental journal·2020
Same author

Evidence for elevated psychiatric distress, poor sleep, and quality of life concerns during the COVID-19 pandemic among U.S. young adults with suspected and reported psychiatric diagnoses.

Psychiatry research·2020

Area of Science:

  • Cognitive Psychology
  • Computational Neuroscience
  • Computer Vision

Background:

  • Symmetry detection is a fundamental visual process.
  • Human perception of symmetry varies based on axis orientation.
  • Artificial neural networks offer a model for understanding visual processing.

Purpose of the Study:

  • To investigate human and artificial neural network performance in detecting symmetry in binary patterns.
  • To compare the influence of training on symmetry detection speed and accuracy.
  • To explore the role of early visual experience in symmetry perception.

Main Methods:

  • Human subjects identified symmetry axes (vertical, horizontal, oblique) in 6x6 binary patterns.
  • Back-propagation neural networks were trained to categorize patterns by symmetry axis.

Related Experiment Videos

  • Network performance was compared to human detection times using cascaded activation functions.

    • Main Results:

    • Networks trained on vertical and horizontal patterns showed better correspondence with human detection times.
    • Pre-training networks with single-axis bars (vertical, horizontal, oblique) accelerated learning of symmetrical patterns.
    • Faster detection of vertical and horizontal symmetries in humans may stem from greater early exposure to these orientations.

    Conclusions:

    • Artificial neural networks can model aspects of human symmetry detection.
    • Training and prior experience significantly impact the speed and efficiency of symmetry perception.
    • The findings support theories linking early visual experience to preferential processing of vertical and horizontal symmetries.