Strengthening surf descriptor with discriminant image filter learning

Application to face recognition

Hamdi Bouchech, Sebti Foufou, Mongi Abidi

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

    Abstract

    Face recognition in extreme situations is still challenging to researchers. While several algorithms have shown great recognition results in ideal conditions, accuracy decreases when recognition tasks present a high illumination variation. In this paper, we propose to add two components to the recognition system to make the surf descriptor efficient in such extreme situations. First, we learn a discriminant image filter that maximizes the discrimination of surf. Second, the obtained discriminant surf (d-surf) is further strengthened by using multispectral images instead of broad band images. DSURF and multispectral d-surf (MD-SURF) were evaluated against two face databases: the feret database, which served as a benchmark, and the iris-m3 multispectral face database, which presented sun lighted faces. Our algorithms have been evaluated against three state-of-the-art algorithms that are MBLBP, HGPP and LGBPHS. The results validated the superiority of D-SURF over the traditional surf descriptor, while MD-SURF performed best out of all studied algorithms.

    Original languageEnglish (US)
    Title of host publication2014 26th International Conference on Microelectronics, ICM 2014
    PublisherInstitute of Electrical and Electronics Engineers Inc.
    Pages136-139
    Number of pages4
    Volume2015-March
    ISBN (Electronic)9781479981533
    DOIs
    StatePublished - Jan 1 2014
    Event2014 26th International Conference on Microelectronics, ICM 2014 - Doha, Qatar
    Duration: Dec 14 2014Dec 17 2014

    Other

    Other2014 26th International Conference on Microelectronics, ICM 2014
    CountryQatar
    CityDoha
    Period12/14/1412/17/14

    Fingerprint

    Face recognition
    Sun
    Lighting

    Keywords

    • Face
    • FERET
    • filter
    • HGPP
    • illumination
    • IRIS-M
    • LGBPHS
    • MBLBP
    • multispectral
    • SURF

    ASJC Scopus subject areas

    • Electrical and Electronic Engineering

    Cite this

    Bouchech, H., Foufou, S., & Abidi, M. (2014). Strengthening surf descriptor with discriminant image filter learning: Application to face recognition. In 2014 26th International Conference on Microelectronics, ICM 2014 (Vol. 2015-March, pp. 136-139). [7071825] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICM.2014.7071825

    Strengthening surf descriptor with discriminant image filter learning : Application to face recognition. / Bouchech, Hamdi; Foufou, Sebti; Abidi, Mongi.

    2014 26th International Conference on Microelectronics, ICM 2014. Vol. 2015-March Institute of Electrical and Electronics Engineers Inc., 2014. p. 136-139 7071825.

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

    Bouchech, H, Foufou, S & Abidi, M 2014, Strengthening surf descriptor with discriminant image filter learning: Application to face recognition. in 2014 26th International Conference on Microelectronics, ICM 2014. vol. 2015-March, 7071825, Institute of Electrical and Electronics Engineers Inc., pp. 136-139, 2014 26th International Conference on Microelectronics, ICM 2014, Doha, Qatar, 12/14/14. https://doi.org/10.1109/ICM.2014.7071825
    Bouchech H, Foufou S, Abidi M. Strengthening surf descriptor with discriminant image filter learning: Application to face recognition. In 2014 26th International Conference on Microelectronics, ICM 2014. Vol. 2015-March. Institute of Electrical and Electronics Engineers Inc. 2014. p. 136-139. 7071825 https://doi.org/10.1109/ICM.2014.7071825
    Bouchech, Hamdi ; Foufou, Sebti ; Abidi, Mongi. / Strengthening surf descriptor with discriminant image filter learning : Application to face recognition. 2014 26th International Conference on Microelectronics, ICM 2014. Vol. 2015-March Institute of Electrical and Electronics Engineers Inc., 2014. pp. 136-139
    @inproceedings{64f49ebce2aa4045b078f9fc8b4eb4f8,
    title = "Strengthening surf descriptor with discriminant image filter learning: Application to face recognition",
    abstract = "Face recognition in extreme situations is still challenging to researchers. While several algorithms have shown great recognition results in ideal conditions, accuracy decreases when recognition tasks present a high illumination variation. In this paper, we propose to add two components to the recognition system to make the surf descriptor efficient in such extreme situations. First, we learn a discriminant image filter that maximizes the discrimination of surf. Second, the obtained discriminant surf (d-surf) is further strengthened by using multispectral images instead of broad band images. DSURF and multispectral d-surf (MD-SURF) were evaluated against two face databases: the feret database, which served as a benchmark, and the iris-m3 multispectral face database, which presented sun lighted faces. Our algorithms have been evaluated against three state-of-the-art algorithms that are MBLBP, HGPP and LGBPHS. The results validated the superiority of D-SURF over the traditional surf descriptor, while MD-SURF performed best out of all studied algorithms.",
    keywords = "Face, FERET, filter, HGPP, illumination, IRIS-M, LGBPHS, MBLBP, multispectral, SURF",
    author = "Hamdi Bouchech and Sebti Foufou and Mongi Abidi",
    year = "2014",
    month = "1",
    day = "1",
    doi = "10.1109/ICM.2014.7071825",
    language = "English (US)",
    volume = "2015-March",
    pages = "136--139",
    booktitle = "2014 26th International Conference on Microelectronics, ICM 2014",
    publisher = "Institute of Electrical and Electronics Engineers Inc.",

    }

    TY - GEN

    T1 - Strengthening surf descriptor with discriminant image filter learning

    T2 - Application to face recognition

    AU - Bouchech, Hamdi

    AU - Foufou, Sebti

    AU - Abidi, Mongi

    PY - 2014/1/1

    Y1 - 2014/1/1

    N2 - Face recognition in extreme situations is still challenging to researchers. While several algorithms have shown great recognition results in ideal conditions, accuracy decreases when recognition tasks present a high illumination variation. In this paper, we propose to add two components to the recognition system to make the surf descriptor efficient in such extreme situations. First, we learn a discriminant image filter that maximizes the discrimination of surf. Second, the obtained discriminant surf (d-surf) is further strengthened by using multispectral images instead of broad band images. DSURF and multispectral d-surf (MD-SURF) were evaluated against two face databases: the feret database, which served as a benchmark, and the iris-m3 multispectral face database, which presented sun lighted faces. Our algorithms have been evaluated against three state-of-the-art algorithms that are MBLBP, HGPP and LGBPHS. The results validated the superiority of D-SURF over the traditional surf descriptor, while MD-SURF performed best out of all studied algorithms.

    AB - Face recognition in extreme situations is still challenging to researchers. While several algorithms have shown great recognition results in ideal conditions, accuracy decreases when recognition tasks present a high illumination variation. In this paper, we propose to add two components to the recognition system to make the surf descriptor efficient in such extreme situations. First, we learn a discriminant image filter that maximizes the discrimination of surf. Second, the obtained discriminant surf (d-surf) is further strengthened by using multispectral images instead of broad band images. DSURF and multispectral d-surf (MD-SURF) were evaluated against two face databases: the feret database, which served as a benchmark, and the iris-m3 multispectral face database, which presented sun lighted faces. Our algorithms have been evaluated against three state-of-the-art algorithms that are MBLBP, HGPP and LGBPHS. The results validated the superiority of D-SURF over the traditional surf descriptor, while MD-SURF performed best out of all studied algorithms.

    KW - Face

    KW - FERET

    KW - filter

    KW - HGPP

    KW - illumination

    KW - IRIS-M

    KW - LGBPHS

    KW - MBLBP

    KW - multispectral

    KW - SURF

    UR - http://www.scopus.com/inward/record.url?scp=84940758191&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84940758191&partnerID=8YFLogxK

    U2 - 10.1109/ICM.2014.7071825

    DO - 10.1109/ICM.2014.7071825

    M3 - Conference contribution

    VL - 2015-March

    SP - 136

    EP - 139

    BT - 2014 26th International Conference on Microelectronics, ICM 2014

    PB - Institute of Electrical and Electronics Engineers Inc.

    ER -