Overview of the proposed method. MagFace learns for (a) in-the-wild faces (b) a universal embedding by pulling the easier samples closer to the class center and pushing them away from the origin o. The magnitude l before normalization increases along with feature's cosine distance to its class center, and therefore reveals the quality for each face. The larger the l, the more likely the sample can be recognized.

People

Qiang Meng, Shichao Zhao, Zhida Huang, Feng Zhou

Abstract

The performance of face recognition system degrades when the variability of the acquired faces increases. Prior work alleviates this issue by either monitoring the face quality in pre-processing or predicting the data uncertainty along with the face feature. This paper proposes MagFace, a category of losses that learn a universal feature embedding whose magnitude can measure the quality of the given face. Under the new loss, it can be proven that the magnitude of the feature embedding monotonically increases if the subject is more likely to be recognized. In addition, MagFace introduces an adaptive mechanism to learn a well-structured within-class feature distributions by pulling easy samples to class centers while pushing hard samples away. This prevents models from overfitting on noisy low-quality samples and improves face recognition in the wild. Extensive experiments conducted on face recognition, quality assessments as well as clustering demonstrate its superiority over state-of-the-arts.

Code

Avaiable at here.

Paper

	MagFace: A universal representation for face recognition and quality assessment Qiang Meng, Shichao Zhao, Zhida Huang, Feng Zhou Algorithm Research, Aibee Inc. In Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition [pdf] [supp] [arXiv] [Poster] [Video on BiliBili]

Video

BibTeX

@inproceedings{meng2021magface,
    title = {{MagFace}: A universal representation for face recognition and quality assessment},
    author = {Meng, Qiang and Zhao, Shichao and Huang, Zhida and Zhou, Feng},
    booktitle = CVPR,
    year = 2021
}