ML Research Wiki / Benchmarks / Adversarial Attack / CIFAR-10

CIFAR-10

Adversarial Attack Benchmark

Performance Over Time

📊 Showing 6 results | 📏 Metric: Attack: PGD20

Top Performing Models

Rank Model Paper Attack: PGD20 Date Code
1 Xu et al. An Orthogonal Classifier for Improving the Adversarial Robustness of Neural Networks 78.68 2021-05-19 📦 MTandHJ/roboc
2 3-ensemble of multi-resolution self-ensembles Ensemble everything everywhere: Multi-scale aggregation for adversarial robustness 78.13 2024-08-08 📦 stanislavfort/ensemble-everything-everywhere 📦 ETH-DISCO/self-ensembling
3 TRADES-ANCRA/ResNet18 Enhancing Robust Representation in Adversarial Training: Alignment and Exclusion Criteria 59.70 2023-10-05 📦 changzhang777/ancra
4 AdvTraining [madry2018] Towards Deep Learning Models Resistant to Adversarial Attacks 48.44 2017-06-19 📦 cleverhans-lab/cleverhans 📦 openai/cleverhans 📦 tensorflow/cleverhans
5 TRADES [zhang2019b] Theoretically Principled Trade-off between Robustness and Accuracy 45.90 2019-01-24 📦 yaodongyu/TRADES 📦 zjfheart/Friendly-Adversarial-Training 📦 goldblum/AdversariallyRobustDistillation
6 XU-Net Attention Masks Help Adversarial Attacks to Bypass Safety Detectors 0.00 2024-11-07 📦 FrankShi9/Attention-Mask-Attack

All Papers (6)

MTandHJ/roboc
stanislavfort/ensemble-everything-everywhere ETH-DISCO/self-ensembling
changzhang777/ancra
cleverhans-lab/cleverhans openai/cleverhans
yaodongyu/TRADES zjfheart/Friendly-Adversarial-Training
FrankShi9/Attention-Mask-Attack