こんにちは、futabatoです。

今回は、ArcFace: Additive Angular Margin Loss for Deep Face RecognitionrcFace ( Deng, Jiankang, et al. , 2019 )の論文に目を通したので、論文メモとしてBlogに残しておきます。

ArcFace: Additive Angular Margin Loss for Deep Face RecognitionrcFace

論文の概要

著者: Deng, Jiankang, et al.
年度: 2019
論文URL: https://openaccess.thecvf.com/content_CVPR_2019/html/Deng_ArcFace_Additive_Angular_Margin_Loss_for_Deep_Face_Recognition_CVPR_2019_paper.html
被引用数: 2775
タグ: Target

ArcFace(Additive Angular Margin Loss)です。
Deep Metric Learningの代表的なモデルとなっています。
先日終了したKaggleのHappyWhaleコンペでArcFaceが多く採用されていたのが印象的でした。

HappyWhale ArcFace Baseline (TPU) | Kaggle

Figure 1. Based on the centre and feature normalisation, all identities are distributed on a hypersphere.

ArcFace: Additive Angular Margin Loss for Deep Face RecognitionrcFace

Abstract

One of the main challenges in feature learning using Deep Convolutional Neural Networks (DCNNs) for large-scale face recognition is the design of appropriate loss functions that can enhance the discriminative power. Centre loss penalises the distance between deep features and their corresponding class centres in the Euclidean space to achieve intra-class compactness. SphereFace assumes that the linear transformation matrix in the last fully connected layer can be used as a representation of the class centres in the angular space and therefore penalises the angles between deep features and their corresponding weights in a multiplicative way. Recently, a popular line of research is to incorporate margins in well-established loss functions in order to maximise face class separability. In this paper, we propose an Additive Angular Margin Loss (ArcFace) to obtain highly discriminative features for face recognition. The proposed ArcFace has a clear geometric interpretation due to its exact correspondence to geodesic distance on a hypersphere. We present arguably the most extensive experimental evaluation against all recent state-of-the-art face recognition methods on ten face recognition benchmarks which includes a new large-scale image database with trillions of pairs and a large-scale video dataset. We show that ArcFace consistently outperforms the state of the art and can be easily implemented with negligible computational overhead. To facilitate future research, the code has been made available.