# ArcFace: Additive Angular Margin Loss for Deep Face Recognition

arXiv: https://arxiv.org/abs/1801.07698 · HTML: https://mlsanity.com/paper/1801.07698/

## Key facts

- Cited by 56 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 23 Jan 2018.

## Built on by (most cited first, 30 of 56)

- [FSCE: Few-Shot Object Detection via Contrastive Proposal Encoding](https://mlsanity.com/paper/2103.05950/) (2021)
- [A Survey on Long-Tailed Visual Recognition](https://mlsanity.com/paper/2205.13775/) (2022)
- [LCM-Lookahead for Encoder-based Text-to-Image Personalization](https://mlsanity.com/paper/2404.03620/) (2024)
- [FreqBlender: Enhancing DeepFake Detection by Blending Frequency Knowledge](https://mlsanity.com/paper/2404.13872/) (2024)
- [Fairness-aware Adversarial Perturbation Towards Bias Mitigation for Deployed Deep Models](https://mlsanity.com/paper/2203.01584/) (2022)
- [Text and Image Guided 3D Avatar Generation and Manipulation](https://mlsanity.com/paper/2202.06079/) (2022)
- [Stable-Hair: Real-World Hair Transfer via Diffusion Model](https://mlsanity.com/paper/2407.14078/) (2024)
- [CLIPVG: Text-Guided Image Manipulation Using Differentiable Vector Graphics](https://mlsanity.com/paper/2212.02122/) (2022)
- [SurrogatePrompt: Bypassing the Safety Filter of Text-to-Image Models via Substitution](https://mlsanity.com/paper/2309.14122/) (2023)
- [SYN-MAD 2022: Competition on Face Morphing Attack Detection Based on Privacy-aware Synthetic Training Data](https://mlsanity.com/paper/2208.07337/) (2022)
- [Investigating the Role of Image Retrieval for Visual Localization -- An exhaustive benchmark](https://mlsanity.com/paper/2205.15761/) (2022)
- [FRoundation: Are Foundation Models Ready for Face Recognition?](https://mlsanity.com/paper/2410.23831/) (2024)
- [MagiCapture: High-Resolution Multi-Concept Portrait Customization](https://mlsanity.com/paper/2309.06895/) (2023)
- [Integrating Language Guidance into Vision-based Deep Metric Learning](https://mlsanity.com/paper/2203.08543/) (2022)
- [Considerations on the Evaluation of Biometric Quality Assessment Algorithms](https://mlsanity.com/paper/2303.13294/) (2023)
- [What Should Be Balanced in a "Balanced" Face Recognition Dataset?](https://mlsanity.com/paper/2304.09818/) (2023)
- [Identity-Preserving Aging of Face Images via Latent Diffusion Models](https://mlsanity.com/paper/2307.08585/) (2023)
- [Beyond Alignment: Blind Video Face Restoration via Parsing-Guided Temporal-Coherent Transformer](https://mlsanity.com/paper/2404.13640/) (2024)
- [The 2021 NIST Speaker Recognition Evaluation](https://mlsanity.com/paper/2204.10242/) (2022)
- [Unsupervised Manga Character Re-identification via Face-body and Spatial-temporal Associated Clustering](https://mlsanity.com/paper/2204.04621/) (2022)
- [MorDeephy: Face Morphing Detection Via Fused Classification](https://mlsanity.com/paper/2208.03110/) (2022)
- [FacialFlowNet: Advancing Facial Optical Flow Estimation with a Diverse Dataset and a Decomposed Model](https://mlsanity.com/paper/2409.05396/) (2024)
- [Zero-Shot Dynamic Concept Personalization with Grid-Based LoRA](https://mlsanity.com/paper/2507.17963/) (2025)
- [STC speaker recognition systems for the NIST SRE 2021](https://mlsanity.com/paper/2111.02298/) (2021)
- [SDGMNet: Statistic-based Dynamic Gradient Modulation for Local Descriptor Learning](https://mlsanity.com/paper/2106.04434/) (2021)
- [Grouped Knowledge Distillation for Deep Face Recognition](https://mlsanity.com/paper/2304.04462/) (2023)
- [End-to-End Joint Target and Non-Target Speakers ASR](https://mlsanity.com/paper/2306.02273/) (2023)
- [A novel Facial Recognition technique with Focusing on Masked Faces](https://mlsanity.com/paper/2501.04444/) (2025)
- [NxtPost: User to Post Recommendations in Facebook Groups](https://mlsanity.com/paper/2202.03645/) (2022)
- [A comparison of different atmospheric turbulence simulation methods for image restoration](https://mlsanity.com/paper/2204.08974/) (2022)

## Abstract (arXiv)

Recently, a popular line of research in face recognition is adopting margins in the well-established softmax loss function to maximize class separability. In this paper, we first introduce an Additive Angular Margin Loss (ArcFace), which not only has a clear geometric interpretation but also significantly enhances the discriminative power. Since ArcFace is susceptible to the massive label noise, we further propose sub-center ArcFace, in which each class contains K sub-centers and training samples only need to be close to any of the K positive sub-centers. Sub-center ArcFace encourages one dominant sub-class that contains the majority of clean faces and non-dominant sub-classes that include hard or noisy faces. Based on this self-propelled isolation, we boost the performance through automatically purifying raw web faces under massive real-world noise. Besides discriminative feature embedding, we also explore the inverse problem, mapping feature vectors to face images. Without training any additional generator or discriminator, the pre-trained ArcFace model can generate identity-preserved face images for both subjects inside and outside the training data only by using the network gradient and Batch Normalization (BN) priors. Extensive experiments demonstrate that ArcFace can enhance the discriminative feature embedding as well as strengthen the generative face synthesis.