TransFace: Calibrating transformer training for face recognition from a data-centric perspective
Author(s)
Type
Conference Paper
Abstract
Vision Transformers (ViTs) have demonstrated powerful representation ability in various visual tasks thanks to their intrinsic data-hungry nature. However, we unexpectedly find that ViTs perform vulnerably when applied to face recognition (FR) scenarios with extremely large datasets. We investigate the reasons for this phenomenon and discover that the existing data augmentation approach and hard sample mining strategy are incompatible with ViTs-based FR backbone due to the lack of tailored consideration on preserving face structural information and leveraging each local token information. To remedy these problems, this paper proposes a superior FR model called TransFace, which employs a patch-level data augmentation strategy named DPAP and a hard sample mining strategy named EHSM. Specially, DPAP randomly perturbs the amplitude information of dominant patches to expand sample diversity, which effectively alleviates the overfitting problem in ViTs. EHSM utilizes the information entropy in the local tokens to dynamically adjust the importance weight of easy and hard samples during training, leading to a more stable prediction. Experiments on several benchmarks demonstrate the superiority of our TransFace. Code and models are available at https://github.com/DanJun6737/TransFace.
Date Issued
2024-01-15
Date Acceptance
2023-10-02
Citation
2023 IEEE/CVF International Conference on Computer Vision (ICCV), 2024, pp.20585-20596
ISSN
1550-5499
Publisher
IEEE Computer Society
Start Page
20585
End Page
20596
Journal / Book Title
2023 IEEE/CVF International Conference on Computer Vision (ICCV)
Copyright Statement
Copyright © 2024 IEEE. This ICCV paper is the Open Access version, provided by the Computer Vision Foundation. Except for this watermark, it is identical to the accepted version; the final published version of the proceedings is available on IEEE Xplore.
Source
IEEE/CVF International Conference on Computer Vision (ICCV)
Subjects
Computer Science
Computer Science, Artificial Intelligence
Computer Science, Theory & Methods
Imaging Science & Photographic Technology
Science & Technology
Technology
Publication Status
Published
Start Date
2023-10-02
Finish Date
2023-10-06
Coverage Spatial
Paris, France