Image Manipulation Localization on RealisticTampering
39.2Loc. F1FRAME
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| FRAMESelection=Learned, Fusion=Learned2026.05 | 39.2 | 28.4 | |
| TruFor2026.05 | 36.4 | 26.8 | |
| CAT-Net2026.05 | 33.8 | 24.4 | |
| XGB-Ensemble (pyIFD)Fusion=XGB-Ensemble2026.05 | 31.4 | 22.4 | |
| ManTraNet2026.05 | 31.2 | 22.4 | |
| RF-Ensemble (pyIFD)Fusion=RF-Ensemble2026.05 | 30.1 | 21.3 | |
| Heuristic-K + uniformSelection=Heuristic, Fusion=Uniform2026.05 | 27.4 | 19.2 | |
| Best single pyIFDSelection=Best single2026.05 | 26.1 | 18.1 | |
| Random-K + uniformSelection=Random, Fusion=Uniform2026.05 | 23.7 | 16.1 | |
| Uniform-all pyIFDSelection=None (all), Fusion=Uniform2026.05 | 22.4 | 15.2 | |
| MMFusion2026.05 | 18.7 | 12.8 |