| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| ImageNet-1K (val) | ASR100 | 57 | 4d ago | ||
| CIFAR-10 (test) | BIM | ASR100 | 57 | 4d ago | |
| ImageNet (test) | TGR | ASR (Inc-v3)72.1 | 26 | 4d ago | |
| ImageNet-compatible | MI-FGSM | Inc-v3 Performance100 | 24 | 4d ago | |
| ImageNet 1000 images (val) | SA | Clean Accuracy77.6 | 24 | 4d ago | |
| CIFAR-100 (test) | AA linf | ASR100 | 9 | 4d ago | |
| NTU | BASAR | Attack Success Rate100 | 9 | 4d ago | |
| HDM05 | BASAR | Attack Success Rate10,000 | 9 | 4d ago | |
| ImageNet-compatible (test) | NI | Acc (Inc-v3)100 | 6 | 4d ago | |
| ImageNet | PGN | ComDefend Robustness93.7 | 6 | 4d ago | |
| ImageNet | MI-DI+RAP | Attack Success Rate (Dense-121)99.9 | 6 | 4d ago | |
| ImageNet | MTDAI+RAP-LS | AT Accuracy (L2)73.7 | 6 | 4d ago | |
| ImageNet-1K (test) | MTDI+RAP-LS | ASR (IncRes-v2)100 | 6 | 4d ago | |
| Chinese Traffic Sign Recognition Database (CTSRD) (test) | I-FGSM | ASR (ResNet34)100 | 5 | 4d ago | |
| LVLM Adversarial Evaluation Set | SGMA | LLaVA55.4 | 5 | 4d ago | |
| ResNext-50 | Ours | Fooling Rate100 | 5 | 4d ago | |
| DenseNet-121 | GFCS | Fooling Rate99.9 | 5 | 4d ago | |
| VGG-19 | GFCS | Fooling Rate100 | 5 | 4d ago | |
| Diabetic Retinopathy | Square attack | Average Queries50 | 4 | 4d ago | |
| CIFAR-10 | PI-Attack | Average Queries99.8 | 4 | 4d ago |
