Face Recognition on BFW (test)
98.41AUROCOracle†
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Oracle†Backbone=ArcFace R100, Demographic labels at inference=true2026.06 | 98.41 | 87.84 | 90.94 | 25 | 1.83 | 1.41 | 2.57 | 96.95 | — | |
| AC-LinearBackbone=ArcFace R100, Demographic labels at inference=false2026.06 | 97.84 | 86.95 | 90.9 | 23 | 1.74 | 1 | 2.66 | 95.69 | 0 | |
| AC-RidgeBackbone=ArcFace R100, Demographic labels at inference=false2026.06 | 97.75 | 87.88 | 90.96 | 28 | 1.9 | 2.87 | 2.74 | 95.9 | 4 | |
| FairCalBackbone=ArcFace R100, Demographic labels at inference=false2026.06 | 97.7 | 87.89 | 90.89 | 25 | 1.57 | 0.64 | 2.6 | 95.44 | 0 | |
| FRAPPEBackbone=ArcFace R100, Demographic labels at inference=false2026.06 | 97.67 | 87.66 | 90.76 | 33 | 2.2 | 0.61 | 2.63 | 95.29 | 3 | |
| BaselineBackbone=ArcFace R100, Demographic labels at inference=false2026.06 | 97.65 | 87.77 | 90.83 | 26 | 1.87 | 11.05 | 7.21 | 95.24 | — | |
| FALCONBackbone=ArcFace R100, Demographic labels at inference=false2026.06 | 97.65 | 87.83 | 90.85 | 25 | 1.8 | 18.73 | 8.37 | 95.24 | 8 | |
| DemoNorm†Backbone=ArcFace R100, Demographic labels at inference=true2026.06 | 97.62 | 87.81 | 90.84 | 18 | 1.05 | 22.22 | 18.69 | 95.24 | — | |
| FSNBackbone=ArcFace R100, Demographic labels at inference=false2026.06 | 97.57 | 87.7 | 90.77 | 22 | 1.49 | 11.19 | 7.1 | 95.02 | 0 | |
| GST†Backbone=ArcFace R100, Demographic labels at inference=true2026.06 | 97.55 | 87.85 | 90.84 | 21 | 1.18 | 0.5 | 2.61 | 95.24 | — | |
| AC-MLPBackbone=ArcFace R100, Demographic labels at inference=false2026.06 | 97.46 | 87.09 | 90.41 | 21 | 1.54 | 2.2 | 2.85 | 95.23 | 0 |