Face recognition on VGG-Face
100Accuracy (f)Original Model
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Original Model2025.03 | 100 | 100 | 97 | 99.22 | — | 100 | |
| Saliency Unlearn2025.03 | 18.5 | 100 | 11 | 97.67 | 91.46 | 3.5 | |
| Boundary Expand2025.03 | 4.25 | 99.94 | 3 | 97.33 | 95.64 | 0 | |
| Bad Teacher2025.03 | 2.75 | 98.5 | 3 | 95.56 | 94.77 | 0 | |
| Learn to Unlearn2025.03 | 2.25 | 96.94 | 1 | 92.33 | 94.13 | 1.25 | |
| Negative Gradient2025.03 | 1.15 | 98.25 | 1 | 93.67 | 94.82 | 1 | |
| Retrain Model2025.03 | 0 | 100 | 0 | 98.89 | 97.94 | 0 | |
| Random Label2025.03 | 0 | 99.97 | 0 | 97.22 | 97.11 | 0 | |
| Boundary Shrink2025.03 | 0 | 100 | 0 | 97.33 | 97.16 | 0 | |
| Influence Unlearn2025.03 | 0 | 99.97 | 0 | 96.11 | 96.55 | 0 | |
| Ours2025.03 | 0 | 100 | 0 | 97.67 | 97.33 | 0 |