BIROn - Birkbeck Institutional Research Online

    CDPM: convolutional deformable part models for semantically aligned person re-identification

    Wang, K. and Ding, C. and Maybank, Stephen and Tao, D. (2019) CDPM: convolutional deformable part models for semantically aligned person re-identification. IEEE Transactions on Image Processing , ISSN 1057-7149. (In Press)

    [img] Text
    CDPM ConvolutionalDeformable.pdf - Author's Accepted Manuscript
    Restricted to Repository staff only until 30 December 2020.

    Download (1MB) | Request a copy

    Abstract

    Part-level representations are essential for robust person re-identification. However, common errors that arise during pedestrian detection frequently result in severe misalignment problems for body parts, which degrade the quality of part representations. Accordingly, to deal with this problem, we propose a novel method named Convolutional Deformable Part Models (CDPM). CDPM works by decoupling the complex part alignment procedure into two easier steps: first, a vertical alignment step detects each body part in the vertical direction, with the help of a multi-task learning model; second, a horizontal refinement step based on attention suppresses the background information around each detected body part. Since these two steps are performed orthogonally and sequentially, the difficulty of part alignment is significantly reduced. In the testing stage, CDPM is able to accurately align flexible body parts without any need for outside information. Extensive experimental results demonstrate the effectiveness of the proposed CDPM for part alignment. Most impressively, CDPM achieves state-of-the-art performance on three large-scale data sets: Market-1501, DukeMTMC-ReID, and CUHK03.

    Metadata

    Item Type: Article
    Additional Information: (c) 2019 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other users, including reprinting/ republishing this material for advertising or promotional purposes, creating new collective works for resale or redistribution to servers or lists, or reuse of any copyrighted components of this work in other works.
    Keyword(s) / Subject(s): Person re-identification, alignment-robust recognition, part-based model, multi-task learning
    School: School of Business, Economics & Informatics > Computer Science and Information Systems
    Depositing User: Stephen Maybank
    Date Deposited: 16 Dec 2019 10:14
    Last Modified: 22 Jul 2020 01:23
    URI: https://eprints.bbk.ac.uk/id/eprint/30275

    Statistics

    Downloads
    Activity Overview
    1Download
    45Hits

    Additional statistics are available via IRStats2.

    Archive Staff Only (login required)

    Edit/View Item Edit/View Item