Jove
Visualize
Contact Us

Related Experiment Videos

Face recognition using IPCA-ICA algorithm.

Issam Dagher1, Rabih Nachar

  • 1Department of Computer Engineering, University of Balamand, PO Box 100, Elkoura, Lebanon. dagheri@balamand.edu.lb

IEEE Transactions on Pattern Analysis and Machine Intelligence
|May 27, 2006
PubMed
Summary

A new fast incremental principal non-Gaussian directions analysis (IPCA-ICA) algorithm efficiently extracts independent components from image data. This method enhances face recognition accuracy by combining principal component analysis (PCA) and independent component analysis (ICA) for improved performance.

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

Combining contour-based and region-based in image segmentation.

F1000Research·2024
Same author

Facial expression recognition using three-stage support vector machines.

Visual computing for industry, biomedicine, and art·2020
See all related articles
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

Area of Science:

  • Computer Vision
  • Machine Learning
  • Signal Processing

Background:

  • Principal Component Analysis (PCA) and Independent Component Analysis (ICA) are established methods for dimensionality reduction and source separation.
  • Traditional PCA and ICA algorithms can be computationally intensive, especially for large datasets like image databases.
  • Efficiently extracting meaningful features from image data is crucial for applications like face recognition.

Purpose of the Study:

  • To introduce a fast incremental principal non-Gaussian directions analysis (IPCA-ICA) algorithm.
  • To develop a covariance-free method for incrementally computing principal components and transforming them into independent directions.
  • To apply the novel algorithm to the face recognition problem for improved accuracy.

Main Methods:

Related Experiment Videos

  • The IPCA-ICA algorithm incrementally computes principal components without estimating the covariance matrix.
  • It transforms principal components into independent directions by maximizing non-Gaussianity.
  • The method merges sequential runs of PCA and ICA algorithms in a real-time fashion.

Main Results:

  • The IPCA-ICA algorithm efficiently extracts independent components from image data.
  • The algorithm demonstrated a high average success rate in face recognition tasks across various databases.
  • Performance was superior compared to other existing algorithms.

Conclusions:

  • The IPCA-ICA algorithm offers an efficient and effective approach for feature extraction in image analysis.
  • Its application in face recognition yields significant improvements in accuracy.
  • The covariance-free and incremental nature makes it suitable for large-scale image datasets.