Face Verification on AgeDB
98.7AccuracyPFC
Evaluation Results
| Method | Links | |
|---|---|---|
| PFCTraining Dataset=WebFace42M, Backbone=ResNet-200, Sampling Ratio=0.32022.03 | 98.7 | |
| PFCTraining Dataset=WebFace42M, Backbone=ViT-L, Sampling Ratio=0.32022.03 | 98.67 | |
| VPL2022.03 | 98.6 | |
| PFCTraining Dataset=WebFace42M, Backbone=ResNet-100, Sampling Ratio=0.32022.03 | 98.6 | |
| VPL-ArcFaceTraining Dataset=MS1MV32026.04 | 98.6 | |
| GIFTrain Set=MS1MV2, Backbone=ResNet-1002025.05 | 98.58 | |
| GIFTrain Set=WebFace4M, Backbone=ResNet-1002025.05 | 98.55 | |
| WebFace42M2022.03 | 98.53 | |
| PFCTraining Dataset=WebFace12M, Backbone=ResNet-100, Sampling Ratio=0.32022.03 | 98.53 | |
| PFCTraining Dataset=WebFace42M, Backbone=ViT-B, Sampling Ratio=0.32022.03 | 98.53 | |
| BroadFace2022.03 | 98.38 | |
| ArcFaceBackbone=R100, Training Dataset=IBUG500K2018.01 | 98.38 | |
| BroadFaceTrain Set=MS1MV2, Backbone=ResNet-1002025.05 | 98.38 | |
| BroadFaceTraining Dataset=MS1MV22026.04 | 98.38 | |
| CurricularFace2018.01 | 98.37 | |
| ElasticFace-Arc+Training Dataset=MS1MV22026.04 | 98.35 | |
| CurricularFaceBackbone=ResNet1002020.04 | 98.32 | |
| CurricularFace2022.03 | 98.32 | |
| CurricularFaceTrain Set=MS1MV2, Backbone=ResNet-1002025.05 | 98.32 | |
| CurricularFaceTraining Dataset=MS1MV22026.04 | 98.32 | |
| Sub-center2022.03 | 98.31 | |
| ArcFace2022.03 | 98.28 | |
| GroupFace2022.03 | 98.28 | |
| ArcFaceTrain Set=MS1MV2, Backbone=ResNet-1002025.05 | 98.28 | |
| GroupFaceTrain Set=MS1MV2, Backbone=ResNet-1002025.05 | 98.28 | |
| ArcFace(m=0.50)Training Dataset=MS1MV22026.04 | 98.28 | |
| ElasticFace-Cos+Training Dataset=MS1MV22026.04 | 98.28 | |
| PFCTraining Dataset=WebFace42M, Backbone=ResNet-100, Sampling Ratio=0.0082022.03 | 98.27 | |
| FunFaceTraining Dataset=MS1MV32026.04 | 98.27 | |
| DCQ2022.03 | 98.23 | |
| ArcFaceBackbone=R100, Training Dataset=MS1MV32018.01 | 98.23 | |
| PartialFC(ArcFace)Training Dataset=MS1MV22026.04 | 98.2 | |
| GIFTrain Set=WebFace4M, Backbone=ViT-base2025.05 | 98.19 | |
| MagFaceTrain Set=MS1MV2, Backbone=ResNet-1002025.05 | 98.17 | |
| MagFaceTraining Dataset=MS1MV22026.04 | 98.17 | |
| AdaFaceTraining Dataset=MS1MV32026.04 | 98.17 | |
| ArcFace2020.04 | 98.15 | |
| WebFace12M2022.03 | 98.15 | |
| FunFaceTraining Dataset=MS1MV22026.04 | 98.13 | |
| CosFace2022.03 | 98.11 | |
| CosFaceTrain Set=MS1MV2, Backbone=ResNet-1002025.05 | 98.11 | |
| CosFace(m=0.35)Training Dataset=MS1MV22026.04 | 98.11 | |
| AdaFace(m=0.4)Training Dataset=MS1MV22026.04 | 98.05 | |
| PartialFC(CosFace)Training Dataset=MS1MV22026.04 | 98.03 | |
| PFCTraining Dataset=WebFace4M, Backbone=ResNet-100, Sampling Ratio=0.32022.03 | 98.01 | |
| MV-Arc-Softmax*Backbone=ResNet1002020.04 | 97.95 | |
| MV-Softmax2022.03 | 97.95 | |
| MV-Softmax2018.01 | 97.95 | |
| MV-SoftMaxTraining Dataset=MS1MV22026.04 | 97.95 | |
| ArcFaceTraining Dataset=WebFace4M2026.04 | 97.95 | |
| PFCTraining Dataset=WebFace12M, Backbone=ResNet-100, Sampling Ratio=0.0132022.03 | 97.93 | |
| ArcFaceTrain Set=WebFace4M, Backbone=ResNet-1002025.05 | 97.93 | |
| FaceGraph2018.01 | 97.92 | |
| VirFace2022.03 | 97.91 | |
| DB2022.03 | 97.9 | |
| AdaFaceTraining Dataset=WebFace4M2026.04 | 97.9 | |
| FunFaceTraining Dataset=WebFace4M2026.04 | 97.9 | |
| Search-Softmax2018.01 | 97.75 | |
| PFCTraining Dataset=WebFace4M, Backbone=ResNet-100, Sampling Ratio=0.042022.03 | 97.52 | |
| AdaFaceTrain Set=WebFace4M, Backbone=ViT-base2025.05 | 97.48 | |
| MC-mini-AMC2022.03 | 97.25 | |
| SST2022.03 | 97.2 | |
| BioMetricNet2022.03 | 96.12 | |
| AFRN2022.03 | 95.35 | |
| CASIAVenue=(Real), Generator=/, Scale (#IDs × #Imgs)=0.49M (10.5K × 47)2026.05 | 94.5 | |
| MorphFaceVenue=CVPR 25, Generator=DM, Scale (#IDs × #Imgs)=1.2M (24K × 50)2026.05 | 93.27 | |
| SteerFaceVenue=(Ours), Generator=DM, Scale (#IDs × #Imgs)=1.2M (24K × 50)2026.05 | 93.12 | |
| SteerFaceVenue=(Ours), Generator=DM, Scale (#IDs × #Imgs)=0.5M (10K × 50)2026.05 | 92.73 | |
| UIFace*Venue=ICLR 25, Generator=DM, Scale (#IDs × #Imgs)=1.2M (10K × 50)2026.05 | 92.51 | |
| Arc2FaceVenue=ECCV 24, Generator=DM, Scale (#IDs × #Imgs)=1.2M (20K × 50, 40K × 5)2026.05 | 92.45 | |
| IDPerturb*Venue=CVPR 26, Generator=DM, Scale (#IDs × #Imgs)=1.2M (24K × 50)2026.05 | 92.16 | |
| MorphFaceVenue=CVPR 25, Generator=DM, Scale (#IDs × #Imgs)=0.5M (10K × 50)2026.05 | 91.8 | |
| CemiFaceVenue=NeurIPS 24, Generator=DM, Scale (#IDs × #Imgs)=0.5M (10K × 50)2026.05 | 91.33 | |
| ID³Training set=CASIA, Backbone=LResNet50-IR2024.09 | 91 | |
| ID3Venue=NeurIPS 24, Generator=DM, Scale (#IDs × #Imgs)=0.5M (10K × 50)2026.05 | 91 | |
| UIFaceVenue=ICLR 25, Generator=DM, Scale (#IDs × #Imgs)=0.5M (10K × 50)2026.05 | 90.95 | |
| IDPerturb*Venue=CVPR 26, Generator=DM, Scale (#IDs × #Imgs)=0.5M (10K × 50)2026.05 | 90.28 | |
| Arc2FaceVenue=ECCV 24, Generator=DM, Scale (#IDs × #Imgs)=0.5M (10K × 50)2026.05 | 90.18 | |
| DCFaceVenue=CVPR 23, Generator=DM, Scale (#IDs × #Imgs)=1.2M (20K × 50, 40K × 5)2026.05 | 90.07 | |
| DCFaceTraining set=FFHQ+CASIA, Backbone=LResNet50-IR2024.09 | 89.7 | |
| DCFaceVenue=CVPR 23, Generator=DM, Scale (#IDs × #Imgs)=0.5M (10K × 50)2026.05 | 89.7 | |
| DRGAN2020.04 | 88 | |
| IDiff-FaceVenue=ICCV 23, Generator=DM, Scale (#IDs × #Imgs)=0.5M (10K × 50)2026.05 | 86.43 | |
| ID³Training set=FFHQ, Backbone=LResNet50-IR2024.09 | 83.78 | |
| DigiFaceVenue=WACV 23, Generator=Rendering, Scale (#IDs × #Imgs)=1.2M (10K × 72, 100K × 5)2026.05 | 81.1 | |
| ExFaceGANVenue=IJCB 23, Generator=GAN, Scale (#IDs × #Imgs)=0.5M (10K × 50)2026.05 | 78.92 | |
| IDiff-FaceTraining set=FFHQ, Backbone=LResNet50-IR2024.09 | 78.4 | |
| DigiFaceVenue=WACV 23, Generator=Rendering, Scale (#IDs × #Imgs)=0.5M (10K × 50)2026.05 | 76.97 | |
| Reinforcement-guided synthetic data generation frameworkData Generation=DM, Training Set=small-scale CASIA-WebFace [54] subset2026.04 | 76.8 | |
| Authentic CASIA-Webface [54] subsetData Generation=Authentic, Training Set=small-scale CASIA-WebFace [54] subset2026.04 | 74.72 | |
| NegFaceDiffData Generation=DM, Training Set=small-scale CASIA-WebFace [54] subset2026.04 | 74.68 | |
| SFace2Venue=BIOM 24, Generator=GAN, Scale (#IDs × #Imgs)=0.6M (10K × 60)2026.05 | 74.37 | |
| SFaceVenue=IJCB 22, Generator=GAN, Scale (#IDs × #Imgs)=0.6M (10K × 60)2026.05 | 71.68 | |
| SFaceTraining set=FFHQ, Backbone=LResNet50-IR2024.09 | 69.87 | |
| DCFaceData Generation=GAN, Training Set=small-scale CASIA-WebFace [54] subset2026.04 | 69.75 | |
| IDnetData Generation=GAN, Training Set=small-scale CASIA-WebFace [54] subset2026.04 | 68.73 | |
| SFace2Data Generation=GAN, Training Set=small-scale CASIA-WebFace [54] subset2026.04 | 68.12 | |
| SFaceData Generation=GAN, Training Set=small-scale CASIA-WebFace [54] subset2026.04 | 66.82 | |
| IDiff-FaceData Generation=DM, Training Set=small-scale CASIA-WebFace [54] subset2026.04 | 66.6 | |
| ID-NetTraining set=FFHQ, Backbone=LResNet50-IR2024.09 | 63.58 |