mlsanityData to 7 Oct 2026

WebFace260M: A Benchmark for Million-Scale Deep Face Recognition

arXiv 2204.10149 ↗First posted 21 Apr 2022cs.CVcited by 41

Not on the map. Why?

Abstract (arXiv)

Face benchmarks empower the research community to train and evaluate high-performance face recognition systems. In this paper, we contribute a new million-scale recognition benchmark, containing uncurated 4M identities/260M faces (WebFace260M) and cleaned 2M identities/42M faces (WebFace42M) training data, as well as an elaborately designed time-constrained evaluation protocol. Firstly, we collect 4M name lists and download 260M faces from the Internet. Then, a Cleaning Automatically utilizing Self-Training (CAST) pipeline is devised to purify the tremendous WebFace260M, which is efficient and scalable. To the best of our knowledge, the cleaned WebFace42M is the largest public face recognition training set and we expect to close the data gap between academia and industry. Referring to practical deployments, Face Recognition Under Inference Time conStraint (FRUITS) protocol and a new test set with rich attributes are constructed. Besides, we gather a large-scale masked face sub-set for biometrics assessment under COVID-19. For a comprehensive evaluation of face matchers, three recognition tasks are performed under standard, masked and unbiased settings, respectively. Equipped with this benchmark, we delve into million-scale face recognition problems. A distributed framework is developed to train face recognition models efficiently without tampering with the performance. Enabled by WebFace42M, we reduce 40% failure rate on the challenging IJB-C set and rank 3rd among 430 entries on NIST-FRVT. Even 10% data (WebFace4M) shows superior performance compared with the public training sets. Furthermore, comprehensive baselines are established under the FRUITS-100/500/1000 milliseconds protocols. The proposed benchmark shows enormous potential on standard, masked and unbiased face recognition scenarios. Our WebFace260M website is https://www.face-benchmark.org.

Key facts. Cited by 41 papers in the map (data through 7 Oct 2026). Builds on 65 of its 112 references that are in the map. First posted on arXiv 21 Apr 2022.

Builds on

65 of 112 references.

References · 65

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
  3. 2014ImageNet Large Scale Visual Recognition Challengecited by 8,441
  4. 2019EfficientNet: Rethinking Model Scaling for Convolutional Neural Networkscited by 6,415
  5. 2017Squeeze-and-Excitation Networkscited by 4,763
  6. 2014Going Deeper with Convolutionscited by 4,231
  7. 2017MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applicationscited by 3,827
  8. 2015FaceNet: A Unified Embedding for Face Recognition and Clusteringcited by 2,833
  9. 2016Aggregated Residual Transformations for Deep Neural Networkscited by 2,540
  10. 2017Accurate, Large Minibatch SGD: Training ImageNet in 1 Hourcited by 1,479
  11. 2019Self-training with Noisy Student improves ImageNet classificationcited by 1,304
  12. 2018CosFace: Large Margin Cosine Loss for Deep Face Recognitioncited by 968
38 more
  1. 2020Designing Network Design Spacescited by 954
  2. 2017VGGFace2: A dataset for recognising faces across pose and agecited by 911
  3. 2017SphereFace: Deep Hypersphere Embedding for Face Recognitioncited by 733
  4. 2016MS-Celeb-1M: A Dataset and Benchmark for Large-Scale Face Recognitioncited by 621
  5. 2014Learning Face Representation from Scratchcited by 451
  6. 2017Residual Attention Network for Image Classificationcited by 385
  7. 2020CurricularFace: Adaptive Curriculum Learning Loss for Deep Face Recognitioncited by 320
  8. 2018Additive Margin Softmax for Face Verificationcited by 275
  9. 2017Cross-Age LFW: A Database for Studying Cross-Age Face Recognition in Unconstrained Environmentscited by 258
  10. 2017NormFace: L2 Hypersphere Embedding for Face Verificationcited by 237
  11. 2019Billion-scale semi-supervised learning for image classificationcited by 232
  12. 2021WebFace260M: A Benchmark Unveiling the Power of Million-Scale Deep Face Recognitioncited by 221
  13. 2015The MegaFace Benchmark: 1 Million Faces for Recognition at Scalecited by 199
  14. 2018MobileFaceNets: Efficient CNNs for Accurate Real-Time Face Verification on Mobile Devicescited by 180
  15. 2019Probabilistic Face Embeddingscited by 171
  16. 2020Partial FC: Training 10 Million Identities on a Single Machinecited by 163
  17. 2014Deep Learning Face Representation by Joint Identification-Verificationcited by 162
  18. 2019Mitigate Bias in Face Recognition using Skewness-Aware Reinforcement Learningcited by 145
  19. 2015A Light CNN for Deep Face Representation with Noisy Labelscited by 126
  20. 2021A Comprehensive Study on Face Recognition Biases Beyond Demographicscited by 86
  21. 2019AdaCos: Adaptively Scaling Cosine Logits for Effectively Learning Deep Face Representationscited by 83
  22. 2020Towards Universal Representation Learning for Deep Face Recognitioncited by 78
  23. 2020Mitigating Face Recognition Bias via Group Adaptive Classifiercited by 67
  24. 2020Masked Face Recognition Dataset and Applicationcited by 66
  25. 2018The Devil of Face Recognition is in the Noisecited by 65
  26. 2015DeepID3: Face Recognition with Very Deep Neural Networkscited by 62
  27. 2019Mis-classified Vector Guided Softmax Loss for Face Recognitioncited by 60
  28. 2014Deeply learned face representations are sparse, selective, and robustcited by 56
  29. 2019Occlusion Robust Face Recognition Based on Mask Learning with PairwiseDifferential Siamese Networkcited by 51
  30. 2019Learning to Cluster Faces on an Affinity Graphcited by 50
  31. 2020Masked Face Recognition for Secure Authenticationcited by 47
  32. 2017Level Playing Field for Million Scale Face Recognitioncited by 43
  33. 2016UMDFaces: An Annotated Face Dataset for Training Deep Networkscited by 42
  34. 2020Learning to Cluster Faces via Confidence and Connectivity Estimationcited by 42
  35. 2021Masked Face Recognition Challenge: The InsightFace Track Reportcited by 41
  36. 2021Consistent Instance False Positive Improves Fairness in Face Recognitioncited by 40
  37. 2020GroupFace: Learning Latent Groups and Constructing Group-based Representations for Face Recognitioncited by 38
  38. 2020Domain Balancing: Face Recognition on Long-Tailed Domainscited by 32

50 most cited of 65 All 65 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 3; 2023: 4; 2024: 11; 2025: 11; 2026: 12.

41 papers.

Sort:

By area:

  1. 2024Vec2Face: Scaling Face Dataset Generation with Loosely Constrained Vectorscited by 15
  2. 2024TopoFR: A Closer Look at Topology Alignment on Face Recognitioncited by 13
  3. 2025LVFace: Progressive Cluster Optimization for Large Vision Models in Face Recognitioncited by 12
  4. 2024Benchmark Data Repositories for Better Benchmarkingcited by 10
  5. 2024A Generative Approach for Wikipedia-Scale Visual Entity Recognitioncited by 9
  6. 2022OCFR 2022: Competition on Occluded Face Recognition From Synthetically Generated Structure-Aware Occlusionscited by 7
  7. 2023CoNAN: Conditional Neural Aggregation Network For Unconstrained Face Feature Fusioncited by 6
  8. 2024Less can be more: representational vs. stereotypical gender bias in facial expression recognitioncited by 5
  9. 2022Perspective Reconstruction of Human Faces by Joint Mesh and Landmark Regressioncited by 4
  10. 2022Physically-Based Face Rendering for NIR-VIS Face Recognitioncited by 3
  11. 2023ProS: Facial Omni-Representation Learning via Prototype-based Self-Distillationcited by 3
  12. 2024Identity Overlap Between Face Recognition Train/Test Data: Causing Optimistic Bias in Accuracy Measurementcited by 3
29 more
  1. 2024Goldilocks Test Sets for Face Verificationcited by 3
  2. 2024Template-based Multi-Domain Face Recognitioncited by 3
  3. 2024Impact of Sunglasses on One-to-Many Facial Identification Accuracycited by 3
  4. 2023Versatile User Identification in Extended Reality using Pretrained Similarity-Learningcited by 2
  5. 2023Boosting Cross-Quality Face Verification using Blind Face Restorationcited by 1
  6. 2024Local and Global Feature Attention Fusion Network for Face Recognitioncited by 1
  7. 2024RepFace: Refining Closed-Set Noise with Progressive Label Correction for Face Recognitioncited by 1
  8. 2025A Rapid Test for Accuracy and Bias of Face Recognition Technologycited by 1
  9. 2025Adaptive and Robust DBSCAN with Multi-agent Reinforcement Learningcited by 1
  10. 2025FingerVeinSyn-5M: A Million-Scale Dataset and Benchmark for Finger Vein Recognitioncited by 1
  11. 2025Vec2Face+ for Face Dataset Generationcited by 1
  12. 2025Balancing Beyond Discrete Categories: Continuous Demographic Labels for Fair Face Recognitioncited by 0
  13. 2025Reliable and Reproducible Demographic Inference for Fairness in Face Analysiscited by 0
  14. 2025Exploring Complementarity and Explainability in CNNs for Periocular Verification Across Acquisition Distancescited by 0
  15. 2025Leveraging Large-Scale Face Datasets for Deep Periocular Recognition via Ocular Croppingcited by 0
  16. 2025Towards classification-based representation learning for place recognition on LiDAR scanscited by 0
  17. 2025Alpha Divergence Losses for Biometric Verificationcited by 0
  18. 2026ATTN-FIQA: Interpretable Attention-based Face Image Quality Assessment with Vision Transformerscited by 0
  19. 2026BIDO: A Biometric Identity Online Authentication Frameworkcited by 0
  20. 2026Scalable, Energy-Efficient Optical-Neural Architecture for Multiplexed Deepfake Video Detectioncited by 0
  21. 2026Sparsity-Inducing Divergence Losses for Biometric Verificationcited by 0
  22. 2026Quantifying Training Membership Information in the Hyperspherical Embedding Geometry of Face Recognition Modelscited by 0
  23. 2026SCOUT: Semantic Concept Discovery for Open-Vocabulary Editing of face Recognition Templatescited by 0
  24. 2026EXPL-FR: Explaining Face Recognition Models via Vision-Language Alignmentcited by 0
  25. 2026FVeinSyn: Synthetic Finger Vein Image Generatorcited by 0
  26. 2026Revisiting Face Recognition for Monozygotic Twins: The Celeb Twins Test Setcited by 0
  27. 2026SynThermFace: Amplifying Limited Paired Data for Visible-Thermal Face Recognition via Synthetic Data Generationcited by 0
  28. 2026Restore What Matters: Lessons from Joint Restoration and Recognitioncited by 0
  29. 2026Benchmarking the Explanatory Quality of Open-Weight Vision-Language Models in Face Recognitioncited by 0