Two-phase linear reconstruction measure-based classification for face recognition

Jianping Gou, Yong Xu, David Zhang, Qirong Mao, Lan Du, Yongzhao Zhan

    Research output: Contribution to journalArticleResearchpeer-review

    Abstract

    In this article we propose several two-phase representation-based classification (RBC) methods that are inspired by the idea of the two-phase test sample sparse representation (TPTSR) method with L2-norm. We first introduce two simple extensions of TPTSR using L1-norm alone and the combination of L1-norm and L2-norm, respectively. We then propose two-phase linear reconstruction measure-based classification (TPLRMC) by adopting the linear reconstruction measure (LRM). Decomposing each feature sample as a weighted linear combination of the other feature samples, TPLRMC can measure the similarities between any pairs of feature samples. The linear reconstruction coefficients can capture the feature's neighborhood structure that is hidden in data. Thus, these coefficients with Lp-norm regularization can be used as good similarity measures between samples and the test ones in classifier design of TPLRMC to enhance discriminative capability. In regard to the classification procedure, TPLRMC first coarsely searches K nearest neighbors for a given query sample with LRM, then finely represents the query sample as a linear combination of the chosen K nearest neighbors, and finally uses LRM to perform classification. The experimental results on six face recognition databases and two object recognition databases demonstrate that the proposed methods outperform the competitors used in the experiments.

    Original languageEnglish
    Pages (from-to)17-36
    Number of pages20
    JournalInformation Sciences
    Volume433-434
    DOIs
    Publication statusPublished - 1 Apr 2018

    Keywords

    • Face recognition
    • Linear reconstruction measure
    • Pattern recognition
    • Representation-based classification
    • Sparse representation

    Cite this

    Gou, Jianping ; Xu, Yong ; Zhang, David ; Mao, Qirong ; Du, Lan ; Zhan, Yongzhao. / Two-phase linear reconstruction measure-based classification for face recognition. In: Information Sciences. 2018 ; Vol. 433-434. pp. 17-36.
    @article{6ed30d02b74641798dc9b3cab29192ad,
    title = "Two-phase linear reconstruction measure-based classification for face recognition",
    abstract = "In this article we propose several two-phase representation-based classification (RBC) methods that are inspired by the idea of the two-phase test sample sparse representation (TPTSR) method with L2-norm. We first introduce two simple extensions of TPTSR using L1-norm alone and the combination of L1-norm and L2-norm, respectively. We then propose two-phase linear reconstruction measure-based classification (TPLRMC) by adopting the linear reconstruction measure (LRM). Decomposing each feature sample as a weighted linear combination of the other feature samples, TPLRMC can measure the similarities between any pairs of feature samples. The linear reconstruction coefficients can capture the feature's neighborhood structure that is hidden in data. Thus, these coefficients with Lp-norm regularization can be used as good similarity measures between samples and the test ones in classifier design of TPLRMC to enhance discriminative capability. In regard to the classification procedure, TPLRMC first coarsely searches K nearest neighbors for a given query sample with LRM, then finely represents the query sample as a linear combination of the chosen K nearest neighbors, and finally uses LRM to perform classification. The experimental results on six face recognition databases and two object recognition databases demonstrate that the proposed methods outperform the competitors used in the experiments.",
    keywords = "Face recognition, Linear reconstruction measure, Pattern recognition, Representation-based classification, Sparse representation",
    author = "Jianping Gou and Yong Xu and David Zhang and Qirong Mao and Lan Du and Yongzhao Zhan",
    year = "2018",
    month = "4",
    day = "1",
    doi = "10.1016/j.ins.2017.12.025",
    language = "English",
    volume = "433-434",
    pages = "17--36",
    journal = "Information Sciences",
    issn = "0020-0255",
    publisher = "Elsevier",

    }

    Two-phase linear reconstruction measure-based classification for face recognition. / Gou, Jianping; Xu, Yong; Zhang, David; Mao, Qirong; Du, Lan; Zhan, Yongzhao.

    In: Information Sciences, Vol. 433-434, 01.04.2018, p. 17-36.

    Research output: Contribution to journalArticleResearchpeer-review

    TY - JOUR

    T1 - Two-phase linear reconstruction measure-based classification for face recognition

    AU - Gou, Jianping

    AU - Xu, Yong

    AU - Zhang, David

    AU - Mao, Qirong

    AU - Du, Lan

    AU - Zhan, Yongzhao

    PY - 2018/4/1

    Y1 - 2018/4/1

    N2 - In this article we propose several two-phase representation-based classification (RBC) methods that are inspired by the idea of the two-phase test sample sparse representation (TPTSR) method with L2-norm. We first introduce two simple extensions of TPTSR using L1-norm alone and the combination of L1-norm and L2-norm, respectively. We then propose two-phase linear reconstruction measure-based classification (TPLRMC) by adopting the linear reconstruction measure (LRM). Decomposing each feature sample as a weighted linear combination of the other feature samples, TPLRMC can measure the similarities between any pairs of feature samples. The linear reconstruction coefficients can capture the feature's neighborhood structure that is hidden in data. Thus, these coefficients with Lp-norm regularization can be used as good similarity measures between samples and the test ones in classifier design of TPLRMC to enhance discriminative capability. In regard to the classification procedure, TPLRMC first coarsely searches K nearest neighbors for a given query sample with LRM, then finely represents the query sample as a linear combination of the chosen K nearest neighbors, and finally uses LRM to perform classification. The experimental results on six face recognition databases and two object recognition databases demonstrate that the proposed methods outperform the competitors used in the experiments.

    AB - In this article we propose several two-phase representation-based classification (RBC) methods that are inspired by the idea of the two-phase test sample sparse representation (TPTSR) method with L2-norm. We first introduce two simple extensions of TPTSR using L1-norm alone and the combination of L1-norm and L2-norm, respectively. We then propose two-phase linear reconstruction measure-based classification (TPLRMC) by adopting the linear reconstruction measure (LRM). Decomposing each feature sample as a weighted linear combination of the other feature samples, TPLRMC can measure the similarities between any pairs of feature samples. The linear reconstruction coefficients can capture the feature's neighborhood structure that is hidden in data. Thus, these coefficients with Lp-norm regularization can be used as good similarity measures between samples and the test ones in classifier design of TPLRMC to enhance discriminative capability. In regard to the classification procedure, TPLRMC first coarsely searches K nearest neighbors for a given query sample with LRM, then finely represents the query sample as a linear combination of the chosen K nearest neighbors, and finally uses LRM to perform classification. The experimental results on six face recognition databases and two object recognition databases demonstrate that the proposed methods outperform the competitors used in the experiments.

    KW - Face recognition

    KW - Linear reconstruction measure

    KW - Pattern recognition

    KW - Representation-based classification

    KW - Sparse representation

    UR - http://www.scopus.com/inward/record.url?scp=85039701606&partnerID=8YFLogxK

    U2 - 10.1016/j.ins.2017.12.025

    DO - 10.1016/j.ins.2017.12.025

    M3 - Article

    VL - 433-434

    SP - 17

    EP - 36

    JO - Information Sciences

    JF - Information Sciences

    SN - 0020-0255

    ER -