Image Classification on ImageNet (Robust Accuracy Evaluation)
49.09Robust AccuracyOurs (5 trees)
Evaluation results
| Method | Links | |
|---|---|---|
| Ours (5 trees)Zero-shot=true, Adversarial fine-tuning=ImageNet, PGD attack steps=20, Perturbation radius (epsilon)=1/255, Number of trees=52026.04 | 49.09 | |
| OursZero-shot=true, Adversarial fine-tuning=ImageNet, PGD attack steps=20, Perturbation radius (epsilon)=1/2552026.04 | 47.81 | |
| AoSZero-shot=true, Adversarial fine-tuning=ImageNet, PGD attack steps=20, Perturbation radius (epsilon)=1/2552026.04 | 47.27 | |
| TeCoAZero-shot=true, Adversarial fine-tuning=ImageNet, PGD attack steps=20, Perturbation radius (epsilon)=1/2552026.04 | 41.48 | |
| PMG-FTZero-shot=true, Adversarial fine-tuning=ImageNet, PGD attack steps=20, Perturbation radius (epsilon)=1/2552026.04 | 38.94 | |
| FAREZero-shot=true, Adversarial fine-tuning=ImageNet, PGD attack steps=20, Perturbation radius (epsilon)=1/2552026.04 | 29.8 | |
| CLIPZero-shot=true, Adversarial fine-tuning=ImageNet, PGD attack steps=20, Perturbation radius (epsilon)=1/2552026.04 | 1.48 |