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

Face recognition using an enhanced independent component analysis approach.

Keun-Chang Kwak1, Witold Pedrycz

  • 1Intelligent Robot Division, Electronics and Telecommunications Research Institute (ETRI), Daejeon 305-350, Korea. kwak@etri.re.kr

IEEE Transactions on Neural Networks
|March 28, 2007
PubMed
Summary

This study introduces Fisher

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

Local Surrogate Models With Residual Fuzzy Rules for Model-Agnostic Explanations.

IEEE transactions on cybernetics·2026
Same author

A Prediction Model Integrating Adaptive-Network-Based Fuzzy Inference System and Fuzzy C-Mean Clustering.

IEEE transactions on cybernetics·2026
Same author

Individual Linguistic Granular Computing: A Granulation-Degranulation-Based Approach.

IEEE transactions on cybernetics·2026
Same author

S<sup>2</sup>FS: Spatially-Aware Separability-Driven Feature Selection in Fuzzy Decision Systems.

IEEE transactions on neural networks and learning systems·2026
Same author

Data-Driven Cation Engineering Guides Electrolyte Design for Sustainable Aqueous Zinc Battery Chemistries.

Advanced materials (Deerfield Beach, Fla.)·2026
Same author

A complex-valued widening spiking neural network.

Neural networks : the official journal of the International Neural Network Society·2026

Area of Science:

  • Computer Vision
  • Machine Learning
  • Pattern Recognition

Background:

  • Independent Component Analysis (ICA) typically uses unsupervised learning and high-order statistics for face representation.
  • Generic ICA methods can be limited in their direct application to discriminative tasks like face recognition.

Purpose of the Study:

  • To enhance Independent Component Analysis (ICA) for improved face recognition.
  • To introduce a novel method, Fisher's ICA (FICA), by combining ICA with Fisher's Linear Discriminant Analysis (LDA).

Main Methods:

  • Developed an enhanced Independent Component Analysis (ICA) method, termed Fisher's ICA (FICA).
  • Integrated Fisher's Linear Discriminant Analysis (LDA) with ICA to create the FICA framework.
  • Conducted comparative analysis using four distance metrics and Support Vector Machines (SVMs) for classification.

Main Results:

  • FICA effectively creates well-separated classes in a low-dimensional subspace.
  • Demonstrated FICA's robustness against significant variations in illumination and facial expressions.
  • Achieved improved classification rates on the Facial Recognition Technology (FERET) database compared to Eigenface, Fisherface, and standard ICA.

Conclusions:

  • The proposed FICA method offers a significant improvement over conventional face recognition techniques.
  • FICA provides a powerful approach for face recognition by leveraging both statistical independence and class separability.
  • The method shows promise for real-world applications requiring robust face identification under varying conditions.

Related Experiment Videos