Eigenvalue correction results in face recognition

    Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademic

    49 Downloads (Pure)


    Eigenvalues of sample covariance matrices are often used in biometrics. It has been known for several decades that even though the sample covariance matrix is an unbiased estimate of the real covariance matrix [Fukunaga,1990], the eigenvalues of the sample covariance matrix are biased estimates of the real eigenvalues [Silverstein,1986]. This bias is particularly dominant when the number of samples used for estimation is in the same order as the number of dimensions, as is often the case in biometrics. We investigate the effects of this bias on error rates in verification experiments and show that eigenvalue correction can improve recognition performance.
    Original languageEnglish
    Title of host publicationProceedings of the 29th Symposium on Information Theory in the Benelux
    Subtitle of host publicationLeuven, Belgium, May 29-30, 2008
    EditorsLiesbet van der Perre, Antoine Dejonghe, Valery Ramon
    PublisherWerkgemeenschap voor Informatie- en Communicatietheorie (WIC)
    Number of pages9
    ISBN (Print)978-90-9023135-8
    Publication statusPublished - May 2008
    Event29th Symposium on Information Theory in the Benelux 2008 - Leuven, Belgium, Leuven, Belgium
    Duration: 29 May 200830 May 2008
    Conference number: 29


    Conference29th Symposium on Information Theory in the Benelux 2008
    Other29-30 May 2008


    • IR-64817
    • EWI-12897
    • SCS-Safety
    • METIS-251018


    Dive into the research topics of 'Eigenvalue correction results in face recognition'. Together they form a unique fingerprint.

    Cite this