AI-generated image detection on DDPM generator (test)
99.29AccuracyNPR
Evaluation results
| Method | Links | ||
|---|---|---|---|
| NPRTraining Dataset=ProGAN2026.01 | 99.29 | 100 | |
| NPRTraining Dataset=ProGAN2026.01 | 95.95 | 99.94 | |
| SAFETraining Dataset=ProGAN2026.01 | 89.4 | 98.25 | |
| SAFETraining Dataset=ProGAN2026.01 | 89.19 | 99.09 | |
| Ours (ViT-L/14)Training Dataset=ProGAN, Backbone=ViT-L/142026.01 | 78.31 | 99.13 | |
| FreDectTraining Dataset=ProGAN2026.01 | 74.15 | 78.38 | |
| Ours (ViT-B/32)Training Dataset=ProGAN, Backbone=ViT-B/322026.01 | 69.41 | 94.74 | |
| Ours (ViT-L/14)Training Dataset=ProGAN, Backbone=ViT-L/142026.01 | 69.21 | 98.42 | |
| UnivFDTraining Dataset=ProGAN2026.01 | 67.32 | 95.79 | |
| FatFormerTraining Dataset=ProGAN2026.01 | 64.69 | 87.26 | |
| Ours (ViT-B/32)Training Dataset=ProGAN, Backbone=ViT-B/322026.01 | 63.68 | 89.72 | |
| FreDectTraining Dataset=ProGAN2026.01 | 62.51 | 64.74 | |
| UnivFDTraining Dataset=ProGAN2026.01 | 62.11 | 92.19 | |
| FatFormerTraining Dataset=ProGAN2026.01 | 55.58 | 63.1 | |
| LGradTraining Dataset=ProGAN2026.01 | 55.17 | 62.56 | |
| LNPTraining Dataset=ProGAN2026.01 | 51.85 | 71.04 | |
| CNNSpotTraining Dataset=ProGAN2026.01 | 50.69 | 74.63 | |
| CNNSpotTraining Dataset=ProGAN2026.01 | 50.48 | 72.6 | |
| GramNetTraining Dataset=ProGAN2026.01 | 50.04 | 67.87 | |
| GramNetTraining Dataset=ProGAN2026.01 | 50.02 | 64.08 | |
| LNPTraining Dataset=ProGAN2026.01 | 49.58 | 40.37 | |
| LGradTraining Dataset=ProGAN2026.01 | 49.41 | 45.24 |