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

An Energy Function and Continuous Edit Process for Graph Matching

Finch1, Wilson, Hancock

  • 1University of York, Department of Computer Science, York, UK, Y01 5DD.

Neural Computation
|September 23, 1998
PubMed
Summary

This study introduces a novel nonquadratic energy function for graph matching and a graduated assignment method that updates data graph connections. This approach improves structural error correction and outperforms existing quadratic methods in graph matching tasks.

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

Exhibition of Hydatigenous Placenta.

Transactions. Edinburgh Obstetrical Society·2018
Same author

Exhibition of Hydrocephalic Fœtus.

Transactions. Edinburgh Obstetrical Society·2018
Same author

Exhibition of Right Ovary.

Transactions. Edinburgh Obstetrical Society·2018
Same author

Transanal excision of early rectal carcinoma-review of a personal series.

Colorectal disease : the official journal of the Association of Coloproctology of Great Britain and Ireland·2013
Same author

Adenocarcinoma arising in an ileostomy managed by resection and an ileoanal J pouch.

Colorectal disease : the official journal of the Association of Coloproctology of Great Britain and Ireland·2013
Same author

Military aspects of early analgesia and anesthesia.

Current researches in anesthesia & analgesia·2010

Area of Science:

  • Computer Vision
  • Machine Learning
  • Graph Theory

Background:

  • Graph matching is crucial for comparing relational structures.
  • Existing methods often rely on quadratic energy functions.
  • Nonquadratic energy functions offer potential for improved accuracy but pose computational challenges.

Purpose of the Study:

  • To develop a new nonquadratic energy function for graph matching.
  • To introduce an improved graduated assignment method for locating matches.
  • To enhance the accuracy and robustness of graph matching algorithms.

Main Methods:

  • A novel nonquadratic energy function is derived from a mixture model using Kullback divergence.
  • The soft-assign ansatz is applied to the derivatives of the new energy function.

Related Experiment Videos

  • A graduated assignment technique is proposed where data graph connection strengths self-update.

    • Main Results:

    • The new energy function is a weighted sum of graph Hamming distances.
    • The proposed method allows for self-updating connection strengths in the data graph.
    • Experimental evaluation demonstrates superior performance compared to quadratic counterparts.

    Conclusions:

    • The developed nonquadratic energy function and graduated assignment method offer significant advancements in graph matching.
    • The ability to update data graph structures rectifies errors and improves matching accuracy.
    • This approach represents a promising direction for complex relational data analysis.