Fake Image Detection on Midjourney
98.3AAJoint
Evaluation results
| Method | Links | ||
|---|---|---|---|
| JointInitial offline model=CLIP ViT-B/322026.01 | 98.3 | -0.54 | |
| three-stage domain continual learning frameworkInitial offline model=CLIP ViT-B/322026.01 | 93.53 | -1.68 | |
| ERInitial offline model=CLIP ViT-B/322026.01 | 83.94 | -1.19 | |
| EWCInitial offline model=CLIP ViT-B/322026.01 | 80.11 | -19.52 | |
| A-GEMInitial offline model=CLIP ViT-B/322026.01 | 78.33 | -13.31 | |
| OSLAInitial offline model=CLIP ViT-B/322026.01 | 77.86 | -18.04 | |
| SIInitial offline model=CLIP ViT-B/322026.01 | 77.57 | -20.46 | |
| iCaRLInitial offline model=CLIP ViT-B/322026.01 | 76.18 | -17.72 | |
| SeqInitial offline model=CLIP ViT-B/322026.01 | 73.82 | -15.49 |