Face Recognition on LFW 1.0 (full set)
5.8FRR @ FARARCFACE + EM-FRR
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| ARCFACE + EM-FRRBackbone=ResNet50, Pre-trained=ArcFace, FAR level=10^-32022.10 | 5.8 | 4.11 | 5.24 | |
| ARCFACEBackbone=ResNet50, Pre-trained=ArcFace, FAR level=10^-32022.10 | 5.9 | 4.17 | 1.81 | |
| ARCFACE + EM-FARBackbone=ResNet50, Pre-trained=ArcFace, FAR level=10^-32022.10 | 7.2 | 9.16 | 1.19 | |
| ARCFACEBackbone=ResNet50, Pre-trained=ArcFace, FAR level=10^-42022.10 | 7.8 | 10.27 | 4.72 | |
| ARCFACE + EM-CBackbone=ResNet50, Pre-trained=ArcFace, FAR level=10^-32022.10 | 8.1 | 5.15 | 1.2 | |
| ARCFACE + EM-FRRBackbone=ResNet50, Pre-trained=ArcFace, FAR level=10^-42022.10 | 10 | 5.89 | 33.65 | |
| ARCFACE + PASS-GBackbone=ResNet50, Pre-trained=ArcFace, FAR level=10^-32022.10 | 10.7 | 5.22 | 2.11 | |
| ARCFACE + EM-FARBackbone=ResNet50, Pre-trained=ArcFace, FAR level=10^-42022.10 | 15.1 | 11.22 | 2.11 | |
| ARCFACE + EM-CBackbone=ResNet50, Pre-trained=ArcFace, FAR level=10^-42022.10 | 16.4 | 9.18 | 2.44 | |
| ARCFACE + PASS-GBackbone=ResNet50, Pre-trained=ArcFace, FAR level=10^-42022.10 | 31.5 | 4.54 | 6.51 |