Breaking the 99% barrier: optimisation of 3D face recognition

    Research output: Contribution to journalArticleAcademicpeer-review

    8 Citations (Scopus)
    4 Downloads (Pure)

    Abstract

    This study presents optimisations to a three-dimensional (3D) face recognition method the authors published in 2011. The optimisations concern handling and estimation of motion from a single 3D image using the symmetry of the face, fine registration by selection of the maximum score for small variations of the registration parameters and efficient training using automatic outlier removal where only part of the classifier is retrained. The optimisations lead to a staggering performance improvement: the verification rate on Face Recognition Grand Challenge (FRGC) v2 data at false accept rate = 0.1% increases from 94.6 to 99.3% and the identification rate increases from 99 to 99.4%. Both are, to the authors' knowledge, the best scores ever published on the FRGC data. In addition, the registration time was reduced from about 2.5 to 0.2–0.6 s and the number of comparisons has increased from about 11 000 to more than 50 000 per second. For slightly decreased performance, even millions of comparisons can be realised. The fast registration means near real-time recognition with 2–5 images is possible. The optimisations are not specific for this method, but can be beneficial for other 3D face recognition methods as well.
    Original languageUndefined
    Pages (from-to)169-177
    Number of pages10
    JournalIET biometrics
    Volume4
    Issue number3
    DOIs
    Publication statusPublished - 2015

    Keywords

    • METIS-312525
    • IR-95850
    • 3D Face recognition
    • 3D image
    • FRGC v2 data
    • false accept rate
    • performance improvement
    • classifier training
    • three-dimensional face recognition optimisation
    • verification rate
    • image registration parameters
    • motion estimation
    • motion handling
    • automatic outlier removal
    • real-time image recognition
    • EWI-25873
    • SCS-Safety

    Cite this

    @article{87b2413808124ab780e347814c99f8dd,
    title = "Breaking the 99{\%} barrier: optimisation of 3D face recognition",
    abstract = "This study presents optimisations to a three-dimensional (3D) face recognition method the authors published in 2011. The optimisations concern handling and estimation of motion from a single 3D image using the symmetry of the face, fine registration by selection of the maximum score for small variations of the registration parameters and efficient training using automatic outlier removal where only part of the classifier is retrained. The optimisations lead to a staggering performance improvement: the verification rate on Face Recognition Grand Challenge (FRGC) v2 data at false accept rate = 0.1{\%} increases from 94.6 to 99.3{\%} and the identification rate increases from 99 to 99.4{\%}. Both are, to the authors' knowledge, the best scores ever published on the FRGC data. In addition, the registration time was reduced from about 2.5 to 0.2–0.6 s and the number of comparisons has increased from about 11 000 to more than 50 000 per second. For slightly decreased performance, even millions of comparisons can be realised. The fast registration means near real-time recognition with 2–5 images is possible. The optimisations are not specific for this method, but can be beneficial for other 3D face recognition methods as well.",
    keywords = "METIS-312525, IR-95850, 3D Face recognition, 3D image, FRGC v2 data, false accept rate, performance improvement, classifier training, three-dimensional face recognition optimisation, verification rate, image registration parameters, motion estimation, motion handling, automatic outlier removal, real-time image recognition, EWI-25873, SCS-Safety",
    author = "Spreeuwers, {Lieuwe Jan}",
    note = "eemcs-eprint-25873",
    year = "2015",
    doi = "10.1049/iet-bmt.2014.0017",
    language = "Undefined",
    volume = "4",
    pages = "169--177",
    journal = "IET biometrics",
    issn = "2047-4938",
    publisher = "The Institution of Engineering and Technology",
    number = "3",

    }

    Breaking the 99% barrier: optimisation of 3D face recognition. / Spreeuwers, Lieuwe Jan.

    In: IET biometrics, Vol. 4, No. 3, 2015, p. 169-177.

    Research output: Contribution to journalArticleAcademicpeer-review

    TY - JOUR

    T1 - Breaking the 99% barrier: optimisation of 3D face recognition

    AU - Spreeuwers, Lieuwe Jan

    N1 - eemcs-eprint-25873

    PY - 2015

    Y1 - 2015

    N2 - This study presents optimisations to a three-dimensional (3D) face recognition method the authors published in 2011. The optimisations concern handling and estimation of motion from a single 3D image using the symmetry of the face, fine registration by selection of the maximum score for small variations of the registration parameters and efficient training using automatic outlier removal where only part of the classifier is retrained. The optimisations lead to a staggering performance improvement: the verification rate on Face Recognition Grand Challenge (FRGC) v2 data at false accept rate = 0.1% increases from 94.6 to 99.3% and the identification rate increases from 99 to 99.4%. Both are, to the authors' knowledge, the best scores ever published on the FRGC data. In addition, the registration time was reduced from about 2.5 to 0.2–0.6 s and the number of comparisons has increased from about 11 000 to more than 50 000 per second. For slightly decreased performance, even millions of comparisons can be realised. The fast registration means near real-time recognition with 2–5 images is possible. The optimisations are not specific for this method, but can be beneficial for other 3D face recognition methods as well.

    AB - This study presents optimisations to a three-dimensional (3D) face recognition method the authors published in 2011. The optimisations concern handling and estimation of motion from a single 3D image using the symmetry of the face, fine registration by selection of the maximum score for small variations of the registration parameters and efficient training using automatic outlier removal where only part of the classifier is retrained. The optimisations lead to a staggering performance improvement: the verification rate on Face Recognition Grand Challenge (FRGC) v2 data at false accept rate = 0.1% increases from 94.6 to 99.3% and the identification rate increases from 99 to 99.4%. Both are, to the authors' knowledge, the best scores ever published on the FRGC data. In addition, the registration time was reduced from about 2.5 to 0.2–0.6 s and the number of comparisons has increased from about 11 000 to more than 50 000 per second. For slightly decreased performance, even millions of comparisons can be realised. The fast registration means near real-time recognition with 2–5 images is possible. The optimisations are not specific for this method, but can be beneficial for other 3D face recognition methods as well.

    KW - METIS-312525

    KW - IR-95850

    KW - 3D Face recognition

    KW - 3D image

    KW - FRGC v2 data

    KW - false accept rate

    KW - performance improvement

    KW - classifier training

    KW - three-dimensional face recognition optimisation

    KW - verification rate

    KW - image registration parameters

    KW - motion estimation

    KW - motion handling

    KW - automatic outlier removal

    KW - real-time image recognition

    KW - EWI-25873

    KW - SCS-Safety

    U2 - 10.1049/iet-bmt.2014.0017

    DO - 10.1049/iet-bmt.2014.0017

    M3 - Article

    VL - 4

    SP - 169

    EP - 177

    JO - IET biometrics

    JF - IET biometrics

    SN - 2047-4938

    IS - 3

    ER -