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

Optimal, unsupervised learning in invariant object recognition.

G Wallis, R Baddeley

    Neural Computation
    |May 15, 1997
    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

    Validation of a combined ELISA to detect IgG, IgA and IgM antibody responses to SARS-CoV-2 in mild or moderate non-hospitalised patients.

    Journal of immunological methods·2021
    Same author

    Measurement of Typhi Vi antibodies can be used to assess adaptive immunity in patients with immunodeficiency.

    Clinical and experimental immunology·2018
    Same author

    Fever in the returning traveller: the importance of sensitivity.

    BMJ case reports·2015
    Same author

    Some Further Remarks on the Use of the Longitudinal Incision Through the Scalp in Diseases of the Brain.

    Provincial medical & surgical journal·2010
    Same author

    Directional control-response relationships for mining equipment.

    Ergonomics·2010
    Same author

    Bowel obstruction in the postpartum period as a result of caecal volvulus around a large uterine leiomyoma.

    Journal of obstetrics and gynaecology : the journal of the Institute of Obstetrics and Gynaecology·2009
    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 a novel method for creating transformation-invariant object representations by linking object views based on presentation order and spatial similarity. The derived optimal learning rule significantly outperforms simple Hebbian learning in simulations.

    Area of Science:

    • Computational Neuroscience
    • Machine Learning
    • Computer Vision

    Background:

    • Establishing transformation-invariant representations is crucial for object recognition.
    • Current methods may not fully leverage temporal and spatial information.
    • Understanding how neural networks learn invariant features is an ongoing challenge.

    Purpose of the Study:

    • To propose and analyze a novel method for learning transformation-invariant object representations.
    • To derive an optimal linear learning rule based on presentation time distributions.
    • To evaluate the effectiveness of this learning rule in a competitive network for character recognition.

    Main Methods:

    • Associating different object views using temporal order and spatial similarity.

    Related Experiment Videos

  • Deriving an optimal linear learning rule under known presentation time distributions.
  • Simulating a competitive neural network for character recognition tasks.

    • Main Results:

    • The proposed learning rule demonstrates superior performance compared to simple Hebbian learning.
    • The derived theory provides accurate quantitative predictions for optimal network parameters.
    • The method successfully establishes transformation-invariant representations.

    Conclusions:

    • The novel learning rule effectively enables the establishment of transformation-invariant object representations.
    • This approach offers a significant improvement over traditional Hebbian learning methods.
    • The findings have implications for developing more robust and efficient object recognition systems.