AI-generated image detection on StarGAN generator (test)
100APGramNet
Evaluation results
| Method | Links | ||
|---|---|---|---|
| GramNetTraining Dataset=ProGAN2026.01 | 100 | 95.05 | |
| LNPTraining Dataset=ProGAN2026.01 | 100 | 99.9 | |
| FatFormerTraining Dataset=ProGAN2026.01 | 100 | 98.12 | |
| SAFETraining Dataset=ProGAN2026.01 | 100 | 99.92 | |
| Ours (ViT-B/32)Training Dataset=ProGAN, Backbone=ViT-B/322026.01 | 100 | 99.37 | |
| Ours (ViT-L/14)Training Dataset=ProGAN, Backbone=ViT-L/142026.01 | 100 | 99.85 | |
| NPRTraining Data=Diffusion-based data (SDv1.4 and VQDM)2026.04 | 100 | — | |
| LGradTraining Dataset=ProGAN2026.01 | 99.98 | 98.1 | |
| NPRTraining Dataset=ProGAN2026.01 | 99.51 | 97.5 | |
| FreDectTraining Dataset=ProGAN2026.01 | 99.49 | 94.62 | |
| UnivFDTraining Dataset=ProGAN2026.01 | 99.37 | 95.75 | |
| CNNSpotTraining Dataset=ProGAN2026.01 | 98.09 | 90.72 | |
| EffortTraining Data=Diffusion-based data (SDv1.4 and VQDM)2026.04 | 97.08 | — | |
| VIB-NetTraining Data=Diffusion-based data (SDv1.4 and VQDM)2026.04 | 96.67 | — | |
| MAFLTraining Data=Diffusion-based data (SDv1.4 and VQDM)2026.04 | 96.22 | — | |
| UnivfdTraining Data=Diffusion-based data (SDv1.4 and VQDM)2026.04 | 86.84 | — | |
| CLIPpingTraining Data=Diffusion-based data (SDv1.4 and VQDM)2026.04 | 80 | — |