BIROn - Birkbeck Institutional Research Online

    Robust face alignment via deep progressive reinitialization and adaptive error-driven learning

    Shao, X. and Xing, J. and Lyu, J. and Zhou, X. and Shi, Y. and Maybank, Stephen (2021) Robust face alignment via deep progressive reinitialization and adaptive error-driven learning. IEEE Transactions on Pattern Analysis and Machine Intelligence , ISSN 0162-8828. (In Press)

    [img]
    Preview
    Text
    FaceAlignmentDPRv3.pdf - Author's Accepted Manuscript

    Download (13MB) | Preview

    Abstract

    Regression-based face alignment involves learning a series of mapping functions to predict the true landmark from an initial estimation of the alignment. Most existing approaches focus on learning efficacious mapping functions from some feature representations to improve performance. The issues related to the initial alignment estimation and the final learning objective, however, receive less attention. This work proposes a deep regression architecture with progressive reinitialization and a new error-driven learning loss function to explicitly address the above two issues. Given an image with a rough face detection result, the full face region is firstly mapped by a supervised spatial transformer network to a normalized form and trained to regress coarse positions of landmarks. Then, different face parts are further respectively reinitialized to their own normalized states, followed by another regression sub-network to refine the landmark positions. To deal with the inconsistent annotations in existing training datasets, we further propose an adaptive landmark-weighted loss function. It dynamically adjusts the importance of different landmarks according to their learning errors during training without depending on any hyper-parameters manually set by trial and error. A high level of robustness to annotation inconsistencies is thus achieved. The whole deep architecture permits training from end to end, and extensive experimental analyses and comparisons demonstrate its effectiveness and efficiency. We will release the source code, trained models, and experimental results upon the publication of this work.

    Metadata

    Item Type: Article
    Keyword(s) / Subject(s): Face Alignment, Regression Model, Deep Architecture, Supervised Spatial Transformer Network, Adaptive Learning
    School: School of Business, Economics & Informatics > Computer Science and Information Systems
    Depositing User: Steve Maybank
    Date Deposited: 19 Apr 2021 10:25
    Last Modified: 11 Jun 2021 13:26
    URI: https://eprints.bbk.ac.uk/id/eprint/43917

    Statistics

    Downloads
    Activity Overview
    67Downloads
    30Hits

    Additional statistics are available via IRStats2.

    Archive Staff Only (login required)

    Edit/View Item Edit/View Item