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

Invariant image classification using triple-correlation-based neural networks.

A Delopoulos1, A Tirakis, S Kollias

  • 1Div. of Comput. Sci., Nat. Tech. Univ. of Athens.

IEEE Transactions on Neural Networks
|January 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

Chromosomal fusions shaped the genome of the greater hornwrack bryozoan (Flustra foliacea) (Linnaeus, 1758).

The Journal of heredity·2026
Same author

Voxel-Based Morphometry-from Hype to Hope. A Study on Hippocampal Atrophy in Mesial Temporal Lobe Epilepsy.

AJNR. American journal of neuroradiology·2020
Same author

Relation of dopamine receptor 2 binding to pain perception in female fibromyalgia patients with and without depression--A [¹¹C] raclopride PET-study.

European neuropsychopharmacology : the journal of the European College of Neuropsychopharmacology·2015
Same author

Abdominal compartment syndrome (ACS) in a severely burned patient.

Annals of burns and fire disasters·2015
Same author

Genome- and transcriptome-assisted development of nuclear insertion/deletion markers for Calanus species (Copepoda: Calanoida) identification.

Molecular ecology resources·2014
Same author

Multiple intracranial meningiomas and cavernous hemangiomas.

The neuroradiology journal·2013
Same journal

Universal perceptron and DNA-like learning algorithm for binary neural networks: LSBF and PBF implementations.

IEEE transactions on neural networks·2013
Same journal

Guest editorial: special section on white box nonlinear prediction models.

IEEE transactions on neural networks·2011
Same journal

Data-based fault-tolerant control of high-speed trains with traction/braking notch nonlinearities and actuator failures.

IEEE transactions on neural networks·2011
Same journal

Guest editorial: special section on data-based control, modeling, and optimization.

IEEE transactions on neural networks·2011
Same journal

Neural network-based multiple robot simultaneous localization and mapping.

IEEE transactions on neural networks·2011
Same journal

Data-driven model-free adaptive control for a class of MIMO nonlinear discrete-time systems.

IEEE transactions on neural networks·2011
See all related articles

New triple-correlation-based neural networks offer invariant classification for 2D grayscale images. This method achieves robustness to distortions and noise for reliable image analysis.

Area of Science:

  • Computer Vision
  • Machine Learning
  • Image Processing

Background:

  • Traditional image classification methods struggle with variations like translation, rotation, and dilation.
  • Developing robust image representations is crucial for accurate analysis in diverse conditions.

Purpose of the Study:

  • Introduce and evaluate triple-correlation-based neural networks for invariant classification of 2D grayscale images.
  • Develop an efficient implementation scheme that is robust to distortions and insensitive to noise.

Main Methods:

  • Utilize third-order correlations of images, clustered in spatial or spectral domains, to create invariant representations.
  • Apply neural network architectures directly to these 2D image representations for classification.
  • Demonstrate third-order neural networks as a specific type of triple-correlation-based network.

Related Experiment Videos

Main Results:

  • Achieved invariant image representation robust to translation, rotation, and dilation.
  • The proposed scheme shows robustness to distortions and insensitivity to additive noise.
  • Successful classification of synthetic and real image data using the developed neural networks.

Conclusions:

  • Triple-correlation-based neural networks provide an effective method for invariant image classification.
  • The proposed approach enhances the reliability of image analysis in the presence of various distortions and noise.
  • This technique offers a promising direction for advanced 2D image recognition tasks.