Face Verification on CFP Frontal-Profile
5.02EERHuman Accuracy
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Human Accuracy2016.04 | 5.02 | 94.57 | 98.92 | |
| DREAMBackbone=ResNet-50, Training Data=MS-Celeb-1M, Strategy=end2end+retrain2018.03 | 6.02 | — | — | |
| DREAMBackbone=ResNet-50, Training Data=MS-Celeb-1M, Strategy=end2end2018.03 | 6.43 | — | — | |
| DREAMBackbone=ResNet-18, Training Data=MS-Celeb-1M, Strategy=end2end+retrain2018.03 | 7.03 | — | — | |
| DREAMBackbone=Center-Loss, Training Data=MS-Celeb-1M, Strategy=end2end+retrain2018.03 | 7.26 | — | — | |
| DREAMBackbone=ResNet-50, Training Data=MS-Celeb-1M, Strategy=stitching2018.03 | 7.29 | — | — | |
| DREAMBackbone=ResNet-18, Training Data=MS-Celeb-1M, Strategy=end2end2018.03 | 7.63 | — | — | |
| DREAMBackbone=ResNet-18, Training Data=MS-Celeb-1M, Strategy=stitching2018.03 | 7.71 | — | — | |
| CDFEBackbone=ResNet-50, Training Data=MS-Celeb-1M2018.03 | 7.71 | — | — | |
| DREAMBackbone=Center-Loss, Training Data=MS-Celeb-1M, Strategy=end2end2018.03 | 7.81 | — | — | |
| DREAMBackbone=Center-Loss, Training Data=MS-Celeb-1M, Strategy=stitching2018.03 | 7.82 | — | — | |
| NaïveBackbone=ResNet-50, Training Data=MS-Celeb-1M2018.03 | 7.89 | — | — | |
| JBBackbone=Center-Loss, Training Data=MS-Celeb-1M2018.03 | 8.29 | — | — | |
| CDFEBackbone=ResNet-18, Training Data=MS-Celeb-1M2018.03 | 8.3 | — | — | |
| JBBackbone=ResNet-18, Training Data=MS-Celeb-1M2018.03 | 8.37 | — | — | |
| NaïveBackbone=ResNet-18, Training Data=MS-Celeb-1M2018.03 | 8.4 | — | — | |
| JBBackbone=ResNet-50, Training Data=MS-Celeb-1M2018.03 | 8.49 | — | — | |
| CDFEBackbone=Center-Loss, Training Data=MS-Celeb-1M2018.03 | 8.49 | — | — | |
| NaïveBackbone=Center-Loss, Training Data=MS-Celeb-1M2018.03 | 8.54 | — | — | |
| CNN2016.04 | 8.85 | 89.17 | 97 | |
| FFBackbone=ResNet-50, Training Data=MS-Celeb-1M2018.03 | 14.26 | — | — | |
| FFBackbone=ResNet-18, Training Data=MS-Celeb-1M2018.03 | 14.4 | — | — | |
| FFBackbone=Center-Loss, Training Data=MS-Celeb-1M2018.03 | 14.53 | — | — | |
| Sengupta et al.2016.04 | 14.97 | 84.91 | 93 | |
| ArcFaceBackbone=R100, Training Dataset=MS1MV32018.01 | — | 98.79 | — | |
| ArcFaceBackbone=R100, Training Dataset=IBUG500K2018.01 | — | 98.87 | — | |
| BaselineTraining Data=4.4M2019.04 | — | 92.78 | — | |
| Center Loss2018.01 | — | 77.48 | — | |
| CurricularFace2018.01 | — | 98.36 | — | |
| DR-GANTraining Data=1M2019.04 | — | 93.41 | — | |
| FaceGraph2018.01 | — | 96.9 | — | |
| MV-Softmax2018.01 | — | 98.28 | — | |
| PFE_fuse+matchTraining Data=4.4M2019.04 | — | 93.34 | — | |
| Prodpoly-ResNet50Model=Prodpoly-ResNet502020.06 | — | 98.986 | — | |
| ResNet50Model=ResNet502020.06 | — | 98.8 | — | |
| Search-Softmax2018.01 | — | 95.64 | — | |
| SFace-10Unsupervised=false, Identities=10,575, Samples per Identity=10, Total Samples=105K2022.11 | — | 68.84 | — | |
| SFace-20Unsupervised=false, Identities=10,575, Samples per Identity=20, Total Samples=211K2022.11 | — | 73.33 | — | |
| SFace-40Unsupervised=false, Identities=10,575, Samples per Identity=40, Total Samples=423K2022.11 | — | 73.1 | — | |
| SFace-60Unsupervised=false, Identities=10,575, Samples per Identity=60, Total Samples=634K2022.11 | — | 73.86 | — | |
| TPETraining Data=0.5M2019.04 | — | 89.17 | — | |
| USynthFaceUnsupervised=true, Identities=100K, Samples per Identity=1, Total Samples=100K2022.11 | — | 78.46 | — | |
| USynthFaceUnsupervised=true, Identities=200K, Samples per Identity=1, Total Samples=200K2022.11 | — | 78.03 | — | |
| USynthFaceUnsupervised=true, Identities=400K, Samples per Identity=1, Total Samples=400K2022.11 | — | 78.56 | — | |
| Yin et al.Training Data=0.5M2019.04 | — | 94.39 | — |