Two-dimensional heteroscedastic linear discriminant analysis for age-group classification

Kazuya Ueki, Teruhide Hayashida, Tetsunori Kobayashi

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

    8 Citations (Scopus)

    Abstract

    This paper presents a novel LDA algorithm named 2DHLDA (2-Dimensional Heteroscedastic Linear Discriminant Analysis). The proposed algorithms are applied on age-group classification using facial images under various lighting conditions. 2DHLDA significantly overcomes the singularity problem, so-called 'Small Sample Size' problem (S3 problem), and the original feature space is split into useful dimensions and nuisance dimensions to reduce the influence of different lighting conditions. A two-phased dimensional reduction step, namely 2DHLDA+LDA, is used in our experiment. Our experimental results show that the new 2DHLDA-based approach improves classification accuracy more than the conventional 1D and 2D-based approaches.

    Original languageEnglish
    Title of host publicationProceedings - International Conference on Pattern Recognition
    Pages585-588
    Number of pages4
    Volume2
    DOIs
    Publication statusPublished - 2006
    Event18th International Conference on Pattern Recognition, ICPR 2006 - Hong Kong
    Duration: 2006 Aug 202006 Aug 24

    Other

    Other18th International Conference on Pattern Recognition, ICPR 2006
    CityHong Kong
    Period06/8/2006/8/24

    Fingerprint

    Discriminant analysis
    Lighting
    Experiments

    ASJC Scopus subject areas

    • Electrical and Electronic Engineering
    • Computer Vision and Pattern Recognition
    • Hardware and Architecture

    Cite this

    Ueki, K., Hayashida, T., & Kobayashi, T. (2006). Two-dimensional heteroscedastic linear discriminant analysis for age-group classification. In Proceedings - International Conference on Pattern Recognition (Vol. 2, pp. 585-588). [1699273] https://doi.org/10.1109/ICPR.2006.1138

    Two-dimensional heteroscedastic linear discriminant analysis for age-group classification. / Ueki, Kazuya; Hayashida, Teruhide; Kobayashi, Tetsunori.

    Proceedings - International Conference on Pattern Recognition. Vol. 2 2006. p. 585-588 1699273.

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

    Ueki, K, Hayashida, T & Kobayashi, T 2006, Two-dimensional heteroscedastic linear discriminant analysis for age-group classification. in Proceedings - International Conference on Pattern Recognition. vol. 2, 1699273, pp. 585-588, 18th International Conference on Pattern Recognition, ICPR 2006, Hong Kong, 06/8/20. https://doi.org/10.1109/ICPR.2006.1138
    Ueki K, Hayashida T, Kobayashi T. Two-dimensional heteroscedastic linear discriminant analysis for age-group classification. In Proceedings - International Conference on Pattern Recognition. Vol. 2. 2006. p. 585-588. 1699273 https://doi.org/10.1109/ICPR.2006.1138
    Ueki, Kazuya ; Hayashida, Teruhide ; Kobayashi, Tetsunori. / Two-dimensional heteroscedastic linear discriminant analysis for age-group classification. Proceedings - International Conference on Pattern Recognition. Vol. 2 2006. pp. 585-588
    @inproceedings{015d50a917144a009df8ec5b49c880e3,
    title = "Two-dimensional heteroscedastic linear discriminant analysis for age-group classification",
    abstract = "This paper presents a novel LDA algorithm named 2DHLDA (2-Dimensional Heteroscedastic Linear Discriminant Analysis). The proposed algorithms are applied on age-group classification using facial images under various lighting conditions. 2DHLDA significantly overcomes the singularity problem, so-called 'Small Sample Size' problem (S3 problem), and the original feature space is split into useful dimensions and nuisance dimensions to reduce the influence of different lighting conditions. A two-phased dimensional reduction step, namely 2DHLDA+LDA, is used in our experiment. Our experimental results show that the new 2DHLDA-based approach improves classification accuracy more than the conventional 1D and 2D-based approaches.",
    author = "Kazuya Ueki and Teruhide Hayashida and Tetsunori Kobayashi",
    year = "2006",
    doi = "10.1109/ICPR.2006.1138",
    language = "English",
    isbn = "0769525210",
    volume = "2",
    pages = "585--588",
    booktitle = "Proceedings - International Conference on Pattern Recognition",

    }

    TY - GEN

    T1 - Two-dimensional heteroscedastic linear discriminant analysis for age-group classification

    AU - Ueki, Kazuya

    AU - Hayashida, Teruhide

    AU - Kobayashi, Tetsunori

    PY - 2006

    Y1 - 2006

    N2 - This paper presents a novel LDA algorithm named 2DHLDA (2-Dimensional Heteroscedastic Linear Discriminant Analysis). The proposed algorithms are applied on age-group classification using facial images under various lighting conditions. 2DHLDA significantly overcomes the singularity problem, so-called 'Small Sample Size' problem (S3 problem), and the original feature space is split into useful dimensions and nuisance dimensions to reduce the influence of different lighting conditions. A two-phased dimensional reduction step, namely 2DHLDA+LDA, is used in our experiment. Our experimental results show that the new 2DHLDA-based approach improves classification accuracy more than the conventional 1D and 2D-based approaches.

    AB - This paper presents a novel LDA algorithm named 2DHLDA (2-Dimensional Heteroscedastic Linear Discriminant Analysis). The proposed algorithms are applied on age-group classification using facial images under various lighting conditions. 2DHLDA significantly overcomes the singularity problem, so-called 'Small Sample Size' problem (S3 problem), and the original feature space is split into useful dimensions and nuisance dimensions to reduce the influence of different lighting conditions. A two-phased dimensional reduction step, namely 2DHLDA+LDA, is used in our experiment. Our experimental results show that the new 2DHLDA-based approach improves classification accuracy more than the conventional 1D and 2D-based approaches.

    UR - http://www.scopus.com/inward/record.url?scp=34047193244&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=34047193244&partnerID=8YFLogxK

    U2 - 10.1109/ICPR.2006.1138

    DO - 10.1109/ICPR.2006.1138

    M3 - Conference contribution

    AN - SCOPUS:34047193244

    SN - 0769525210

    SN - 9780769525211

    VL - 2

    SP - 585

    EP - 588

    BT - Proceedings - International Conference on Pattern Recognition

    ER -