mlsanityData to 14 Oct 2026

MagFace: A Universal Representation for Face Recognition and Quality Assessment

arXiv 2103.06627 ↗First posted 11 Mar 2021cs.CVcited by 334

Not on the map. Why?

Abstract (arXiv)

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 wellstructured 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. The code is available at https://github.com/IrvingMeng/MagFace.

Key facts. Cited by 334 papers in the map (data through 14 Oct 2026). Builds on 24 of its 50 references that are in the map. First posted on arXiv 11 Mar 2021.

Builds on

24 of 50 references.

References · 24

  1. 2015FaceNet: A Unified Embedding for Face Recognition and Clusteringcited by 2,849
  2. 2018CosFace: Large Margin Cosine Loss for Deep Face Recognitioncited by 971
  3. 2017VGGFace2: A dataset for recognising faces across pose and agecited by 914
  4. 2017SphereFace: Deep Hypersphere Embedding for Face Recognitioncited by 737
  5. 2016MS-Celeb-1M: A Dataset and Benchmark for Large-Scale Face Recognitioncited by 624
  6. 2014Learning Face Representation from Scratchcited by 453
  7. 2020Circle Loss: A Unified Perspective of Pair Similarity Optimizationcited by 403
  8. 2018Additive Margin Softmax for Face Verificationcited by 278
  9. 2017Cross-Age LFW: A Database for Studying Cross-Age Face Recognition in Unconstrained Environmentscited by 260
  10. 2017NormFace: L2 Hypersphere Embedding for Face Verificationcited by 237
  11. 2019Probabilistic Face Embeddingscited by 173
  12. 2020Data Uncertainty Learning in Face Recognitioncited by 134
12 more
  1. 2020SER-FIQ: Unsupervised Estimation of Face Image Quality Based on Stochastic Embedding Robustnesscited by 112
  2. 2017Deep Metric Learning with Angular Losscited by 107
  3. 2017L2-constrained Softmax Loss for Discriminative Face Verificationcited by 96
  4. 2019AdaCos: Adaptively Scaling Cosine Logits for Effectively Learning Deep Face Representationscited by 83
  5. 2020Face Image Quality Assessment: A Literature Surveycited by 80
  6. 2015DeepID3: Face Recognition with Very Deep Neural Networkscited by 62
  7. 2019FaceQnet: Quality Assessment for Face Recognition based on Deep Learningcited by 62
  8. 2020Learning Meta Face Recognition in Unseen Domainscited by 61
  9. 2020Learning to Cluster Faces via Confidence and Connectivity Estimationcited by 42
  10. 2020The Elements of End-to-end Deep Face Recognition: A Survey of Recent Advancescited by 32
  11. 2018Support Vector Guided Softmax Loss for Face Recognitioncited by 10
  12. 2017Feature Incay for Representation Regularizationcited by 9

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 26; 2022: 52; 2023: 74; 2024: 80; 2025: 44; 2026: 58.

334 papers.

Sort:

By area:

  1. 2022AdaFace: Quality Adaptive Margin for Face Recognitioncited by 391
  2. 2021ElasticFace: Elastic Margin Loss for Deep Face Recognitioncited by 144
  3. 2022DigiFace-1M: 1 Million Digital Face Images for Face Recognitioncited by 92
  4. 2021CR-FIQA: Face Image Quality Assessment by Learning Sample Relative Classifiabilitycited by 69
  5. 2023Residual Denoising Diffusion Modelscited by 69
  6. 2022Physical Adversarial Attack meets Computer Vision: A Decade Surveycited by 67
  7. 2023Real-Time Radiance Fields for Single-Image Portrait View Synthesiscited by 64
  8. 2023Generalizable Synthetic Image Detection via Language-guided Contrastive Learningcited by 49
  9. 2023GANDiffFace: Controllable Generation of Synthetic Datasets for Face Recognition with Realistic Variationscited by 49
  10. 2021Data Poisoning Won't Save You From Facial Recognitioncited by 45
  11. 2022Voice-Face Homogeneity Tells Deepfakecited by 44
  12. 2024KeyPoint Relative Position Encoding for Face Recognitioncited by 38
38 more
  1. 2021Learning Compatible Embeddingscited by 37
  2. 2021PocketNet: Extreme Lightweight Face Recognition Network using Neural Architecture Search and Multi-Step Knowledge Distillationcited by 36
  3. 2021Cross-Quality LFW: A Database for Analyzing Cross-Resolution Image Face Recognition in Unconstrained Environmentscited by 34
  4. 2023EdgeFace: Efficient Face Recognition Model for Edge Devicescited by 33
  5. 2024It's All About Your Sketch: Democratising Sketch Control in Diffusion Modelscited by 29
  6. 2024SDFR: Synthetic Data for Face Recognition Competitioncited by 28
  7. 2023Understanding the Feature Norm for Out-of-Distribution Detectioncited by 27
  8. 2021Face.evoLVe: A High-Performance Face Recognition Librarycited by 26
  9. 2022Rethinking Bias Mitigation: Fairer Architectures Make for Fairer Face Recognitioncited by 26
  10. 2021QMagFace: Simple and Accurate Quality-Aware Face Recognitioncited by 25
  11. 2021SphereFace Revived: Unifying Hyperspherical Face Recognitioncited by 23
  12. 2023DifFIQA: Face Image Quality Assessment Using Denoising Diffusion Probabilistic Modelscited by 23
  13. 2024FRoundation: Are Foundation Models Ready for Face Recognition?cited by 23
  14. 2022Controllable and Guided Face Synthesis for Unconstrained Face Recognitioncited by 22
  15. 2022Cluster and Aggregate: Face Recognition with Large Probe Setcited by 22
  16. 2023TransFace++: Rethinking the Face Recognition Paradigm with a Focus on Accuracy, Efficiency, and Securitycited by 22
  17. 2023Generating Features with Increased Crop-related Diversity for Few-Shot Object Detectioncited by 21
  18. 2023Label-Only Model Inversion Attacks via Knowledge Transfercited by 21
  19. 2024Real-time 3D-aware Portrait Video Relightingcited by 21
  20. 2024Edit Away and My Face Will not Stay: Personal Biometric Defense against Malicious Generative Editingcited by 21
  21. 2021Adversarial Mask: Real-World Universal Adversarial Attack on Face Recognition Modelcited by 19
  22. 2022Improving Federated Learning Face Recognition via Privacy-Agnostic Clusterscited by 18
  23. 2022Part-based Face Recognition with Vision Transformerscited by 18
  24. 2022FaceQAN: Face Image Quality Assessment Through Adversarial Noise Explorationcited by 18
  25. 2021Balancing Biases and Preserving Privacy on Balanced Faces in the Wildcited by 17
  26. 2023A Quality Aware Sample-to-Sample Comparison for Face Recognitioncited by 17
  27. 2024DSL-FIQA: Assessing Facial Image Quality via Dual-Set Degradation Learning and Landmark-Guided Transformercited by 17
  28. 2021SoK: Anti-Facial Recognition Technologycited by 16
  29. 2023Considerations on the Evaluation of Biometric Quality Assessment Algorithmscited by 16
  30. 2023Kaizen: Practical Self-supervised Continual Learning with Continual Fine-tuningcited by 16
  31. 2023Recognizability Embedding Enhancement for Very Low-Resolution Face Recognition and Quality Estimationcited by 16
  32. 2023Single-Shot Implicit Morphable Faces with Consistent Texture Parameterizationcited by 16
  33. 2021Reinforcement Learning-powered Semantic Communication via Semantic Similaritycited by 15
  34. 2023Logical Consistency and Greater Descriptive Power for Facial Hair Attribute Learningcited by 15
  35. 2024Vec2Face: Scaling Face Dataset Generation with Loosely Constrained Vectorscited by 15
  36. 202550 Years of Automated Face Recognitioncited by 15
  37. 2021Pixel-Level Face Image Quality Assessment for Explainable Face Recognitioncited by 14
  38. 2023Generalizing and Decoupling Neural Collapse via Hyperspherical Uniformity Gapcited by 14

50 most cited of 334 All 334 (JSON)