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

Modeling perceptual learning with multiple interacting elements: a neural network model describing early visual

R Peres1, S Hochstein

  • 1Center for Neural Computation, Hebrew University, Jerusalem, Israel.

Journal of Computational Neuroscience
|December 1, 1994
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

Challenging Spontaneous Quantum Collapse with the XENONnT Dark Matter Detector.

Physical review letters·2026
Same author

WIMP Dark Matter Search Using a 3.1 Tonne-Year Exposure of the XENONnT Experiment.

Physical review letters·2025
Same author

Search for Light Dark Matter in Low-Energy Ionization Signals from XENONnT.

Physical review letters·2025
Same author

First Search for Light Dark Matter in the Neutrino Fog with XENONnT.

Physical review letters·2025
Same author

First Indication of Solar ^{8}B Neutrinos via Coherent Elastic Neutrino-Nucleus Scattering with XENONnT.

Physical review letters·2024
Same author

The XENONnT dark matter experiment.

The European physical journal. C, Particles and fields·2024

This study presents a neural network model for early visual processing, enhancing perceptual learning in pop-out detection tasks. The model demonstrates significant performance improvements through a balanced connection strategy, even with noisy data.

Area of Science:

  • Computational neuroscience
  • Visual perception
  • Machine learning

Background:

  • Perceptual learning is crucial for visual processing, particularly in identifying salient stimuli.
  • Psychophysical experiments on odd element detection tasks provide insights into human visual learning.
  • Early visual cortical areas play a fundamental role in initial visual information processing.

Purpose of the Study:

  • To develop a neural network model simulating an early visual cortical area.
  • To understand perceptual learning mechanisms in odd element detection tasks.
  • To investigate the impact of neural network architecture and learning algorithms on visual learning.

Main Methods:

  • Constructed a neural network with orientation-selective units and hypercolumn structure, mimicking monkey neuron receptive fields.

Related Experiment Videos

  • Employed the Associative reward-penalty (Ar-p) reinforcement learning algorithm, informed by studies on cortical plasticity.
  • Simulated odd element detection tasks to evaluate network performance and learning dynamics.

    • Main Results:

    • Network performance improved significantly with a balanced connection strategy: lateral iso-orientation inhibition and cross-orientation facilitation.
    • The model demonstrated effective learning from chance performance levels.
    • The network showed robustness to significant noise in its response function.

    Conclusions:

    • The developed neural network model effectively captures key aspects of perceptual learning in visual odd element detection.
    • Balanced connectivity between orientation-selective units is critical for efficient learning and performance.
    • The model's success suggests a viable computational framework for studying visual cortical plasticity and learning.