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

A Cortical-type Modular Neural Network for Hypothetical Reasoning.

Tomohiko Masutani1, Hiroshi Tsujino, Edgar Koerner

  • 1Wako Research Center, HONDA R&D Co. Ltd, Japan

Neural Networks : the Official Journal of the International Neural Network Society
|July 1, 1997
PubMed
Summary

This study introduces a novel neural network architecture inspired by the neocortex for hypothetical reasoning. It enables robust face recognition through a refined hypothesis-generation process, paving the way for autonomous learning.

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

Effects of the addition of CeO<sub>2</sub> on the steam reforming of ethanol using novel carbon-Al<sub>2</sub>O<sub>3</sub> and carbon-ZrO<sub>2</sub> composite-supported Co catalysts.

RSC advances·2022
Same author

Real-time simulation of a spiking neural network model of the basal ganglia circuitry using general purpose computing on graphics processing units.

Neural networks : the official journal of the International Neural Network Society·2011
Same author

A model for learning topographically organized parts-based representations of objects in visual cortex: topographic nonnegative matrix factorization.

Neural computation·2009
Same author

[A brain-like computer involving emotion and volition].

Brain and nerve = Shinkei kenkyu no shinpo·2008
Same author

A computational model of monkey cortical grating cells.

Biological cybernetics·2004
Same author

Output-driven operation and memory-based architecture principles embedded in a real-world device.

Journal of integrative neuroscience·2004

Area of Science:

  • Computational Neuroscience
  • Artificial Intelligence
  • Cognitive Science

Background:

  • The human cortex performs complex hypothetical reasoning to interpret sensory information.
  • Existing neural network models often lack the nuanced processing capabilities of the neocortex.
  • Understanding the cortex's functional organization is key to developing advanced AI.

Purpose of the Study:

  • To propose a multilayer neural network architecture capable of hypothetical reasoning.
  • To model the sensory input processing observed in the neocortex.
  • To demonstrate robust recognition capabilities, exemplified by face recognition.

Main Methods:

  • Developed a multilayer neural network using modules inspired by neocortical columns, not single neurons.

Related Experiment Videos

  • Implemented a system where initial hypotheses are refined based on input descriptions and feedback.
  • Separated forward input description from feedback-generated hypotheses to control refinement.

    • Main Results:

    • Successfully simulated face recognition within the proposed neocortical architecture.
    • Demonstrated robust recognition through the separation of input and feedback mechanisms.
    • The architecture shows potential for autonomous learning by utilizing discrepancies in descriptions.

    Conclusions:

    • The proposed architecture effectively mimics hypothetical reasoning in the cortex.
    • This approach offers a pathway to more sophisticated and adaptive artificial intelligence systems.
    • The modular design and refinement process contribute to robust and potentially self-improving recognition.