AI-generated image detection on AIGIBench StyleSwim
95.2AccuracyC2P-CLIP
Evaluation results
| Method | Links | ||
|---|---|---|---|
| C2P-CLIPevaluation_strategy=cross-model2025.01 | 95.2 | 100 | |
| DFFreqevaluation_strategy=cross-model2025.01 | 94.7 | 98.6 | |
| UnivFDevaluation_strategy=cross-model2025.01 | 94.5 | 99.3 | |
| LEGOVenue=-2026.05 | 94.1 | 98.4 | |
| FatFormerevaluation_strategy=cross-model2025.01 | 93.9 | 99.8 | |
| AIDEevaluation_strategy=cross-model2025.01 | 85.2 | 85.1 | |
| NPRevaluation_strategy=cross-model2025.01 | 84.7 | 91.7 | |
| LaDeDaevaluation_strategy=cross-model2025.01 | 82.6 | 86.4 | |
| FreqNetevaluation_strategy=cross-model2025.01 | 80.8 | 91.8 | |
| LGradevaluation_strategy=cross-model2025.01 | 68.1 | 75 | |
| CNN-Spotevaluation_strategy=cross-model2025.01 | 58 | 88.9 |