Combined training strategy for low-resolution face recognition with limited application-specific data

Dan Zeng, Luuk Spreeuwers, Raymond Veldhuis, Qijun Zhao*

*Corresponding author for this work

    Research output: Contribution to journalArticleAcademicpeer-review

    6 Downloads (Pure)

    Abstract

    Application-specific data for certain biometric applications are often not sufficiently available. The authors present a solution for face recognition with limited application-specific data. Existing methods often use a classifier with convolutional neural networks (CNNs) as feature extractors. The CNNs are trained with massive general (i.e. not application specific) data and the classifier is trained with application-specific data. Alternatively, the authors propose a combined training strategy to train the classifier on a balanced mixture of general and application-specific data, such that the recognition performance is maximised. The proposed method largely alleviates the needs for application-specific data. To prove its effectiveness, they apply the proposed method to low-resolution face recognition. Specifically, they use the heterogeneous joint Bayesian (HJB) classifier that is capable of comparing features from the same modality but with different characteristics. To further boost performance, the authors augment the training data by pre-processing it to resemble application-specific data. They conducted extensive experiments on challenging datasets, namely, SCface and COX. The results show that the proposed method improves the true match rate on SCface at a false match rate of 10% by ∼11% and the true match rate on COX at a false match rate of 1% by ∼12%.

    Original languageEnglish
    Pages (from-to)1790-1796
    Number of pages7
    JournalIET Image Processing
    Volume13
    Issue number10
    DOIs
    Publication statusE-pub ahead of print/First online - 9 Sep 2019

    Fingerprint

    Face recognition
    Classifiers
    Neural networks
    Biometrics
    Processing

    Cite this

    @article{be1742a95e7444fd8d99f5ccb4c41030,
    title = "Combined training strategy for low-resolution face recognition with limited application-specific data",
    abstract = "Application-specific data for certain biometric applications are often not sufficiently available. The authors present a solution for face recognition with limited application-specific data. Existing methods often use a classifier with convolutional neural networks (CNNs) as feature extractors. The CNNs are trained with massive general (i.e. not application specific) data and the classifier is trained with application-specific data. Alternatively, the authors propose a combined training strategy to train the classifier on a balanced mixture of general and application-specific data, such that the recognition performance is maximised. The proposed method largely alleviates the needs for application-specific data. To prove its effectiveness, they apply the proposed method to low-resolution face recognition. Specifically, they use the heterogeneous joint Bayesian (HJB) classifier that is capable of comparing features from the same modality but with different characteristics. To further boost performance, the authors augment the training data by pre-processing it to resemble application-specific data. They conducted extensive experiments on challenging datasets, namely, SCface and COX. The results show that the proposed method improves the true match rate on SCface at a false match rate of 10{\%} by ∼11{\%} and the true match rate on COX at a false match rate of 1{\%} by ∼12{\%}.",
    author = "Dan Zeng and Luuk Spreeuwers and Raymond Veldhuis and Qijun Zhao",
    year = "2019",
    month = "9",
    day = "9",
    doi = "10.1049/iet-ipr.2018.5732",
    language = "English",
    volume = "13",
    pages = "1790--1796",
    journal = "IET Image Processing",
    issn = "1751-9659",
    publisher = "Institution of Engineering and Technology",
    number = "10",

    }

    Combined training strategy for low-resolution face recognition with limited application-specific data. / Zeng, Dan; Spreeuwers, Luuk; Veldhuis, Raymond; Zhao, Qijun.

    In: IET Image Processing, Vol. 13, No. 10, 09.09.2019, p. 1790-1796.

    Research output: Contribution to journalArticleAcademicpeer-review

    TY - JOUR

    T1 - Combined training strategy for low-resolution face recognition with limited application-specific data

    AU - Zeng, Dan

    AU - Spreeuwers, Luuk

    AU - Veldhuis, Raymond

    AU - Zhao, Qijun

    PY - 2019/9/9

    Y1 - 2019/9/9

    N2 - Application-specific data for certain biometric applications are often not sufficiently available. The authors present a solution for face recognition with limited application-specific data. Existing methods often use a classifier with convolutional neural networks (CNNs) as feature extractors. The CNNs are trained with massive general (i.e. not application specific) data and the classifier is trained with application-specific data. Alternatively, the authors propose a combined training strategy to train the classifier on a balanced mixture of general and application-specific data, such that the recognition performance is maximised. The proposed method largely alleviates the needs for application-specific data. To prove its effectiveness, they apply the proposed method to low-resolution face recognition. Specifically, they use the heterogeneous joint Bayesian (HJB) classifier that is capable of comparing features from the same modality but with different characteristics. To further boost performance, the authors augment the training data by pre-processing it to resemble application-specific data. They conducted extensive experiments on challenging datasets, namely, SCface and COX. The results show that the proposed method improves the true match rate on SCface at a false match rate of 10% by ∼11% and the true match rate on COX at a false match rate of 1% by ∼12%.

    AB - Application-specific data for certain biometric applications are often not sufficiently available. The authors present a solution for face recognition with limited application-specific data. Existing methods often use a classifier with convolutional neural networks (CNNs) as feature extractors. The CNNs are trained with massive general (i.e. not application specific) data and the classifier is trained with application-specific data. Alternatively, the authors propose a combined training strategy to train the classifier on a balanced mixture of general and application-specific data, such that the recognition performance is maximised. The proposed method largely alleviates the needs for application-specific data. To prove its effectiveness, they apply the proposed method to low-resolution face recognition. Specifically, they use the heterogeneous joint Bayesian (HJB) classifier that is capable of comparing features from the same modality but with different characteristics. To further boost performance, the authors augment the training data by pre-processing it to resemble application-specific data. They conducted extensive experiments on challenging datasets, namely, SCface and COX. The results show that the proposed method improves the true match rate on SCface at a false match rate of 10% by ∼11% and the true match rate on COX at a false match rate of 1% by ∼12%.

    U2 - 10.1049/iet-ipr.2018.5732

    DO - 10.1049/iet-ipr.2018.5732

    M3 - Article

    AN - SCOPUS:85072018331

    VL - 13

    SP - 1790

    EP - 1796

    JO - IET Image Processing

    JF - IET Image Processing

    SN - 1751-9659

    IS - 10

    ER -