paper-with-me

홈 › Papers

Improved Image Wasserstein Attacks and Defenses

2020-04-26 · Edward J. Hu, Adith Swaminathan, Hadi Salman, Greg Yang

Robustness against image perturbations bounded by a $\ell_p$ ball have been well-studied in recent literature. Perturbations in the real-world, however, rarely exhibit the pixel independence that $\ell_p$ threat models assume. A recently proposed Wasserstein distance-bounded threat model is a promising alternative that limits the perturbation to pixel mass movements. We point out and rectify flaws in previous definition of the Wasserstein threat model and explore stronger attacks and defenses under our better-defined framework. Lastly, we discuss the inability of current Wasserstein-robust models in defending against perturbations seen in the real world. Our code and trained models are available at https://github.com/edwardjhu/improved_wasserstein .

📄 PDF Abstract BibTeX arXiv:2004.12478

Code (1)

edwardjhu/improved_wasserstein 공식 구현 pytorch

Similar Papers 제목 키워드 기반

Wasserstein Smoothing: Certified Robustness against Wasserstein Adversarial Attacks

2019-10-23 · Alexander Levine, Soheil Feizi

In the last couple of years, several adversarial attack methods based on different threat models have been proposed for the image classification problem. Most existing defenses consider additive threat models in which sa…

Adversarial Attackimage-classificationImage Classification

Guided Adversarial Attack for Evaluating and Enhancing Adversarial Defenses

2020-11-30 · NeurIPS 2020 12 · Gaurang Sriramanan, Sravanti Addepalli, Arya Baburaj, R. Venkatesh Babu

Advances in the development of adversarial attacks have been fundamental to the progress of adversarial defense research. Efficient and effective attacks are crucial for reliable evaluation of defenses, and also for deve…

Adversarial AttackAdversarial Defense

MARS: A Malignity-Aware Backdoor Defense in Federated Learning

2025-09-21 · Wei Wan, Yuxuan Ning, Zhicong Huang, Cheng Hong 외 arxiv

Federated Learning (FL) is a distributed paradigm aimed at protecting participant data privacy by exchanging model parameters to achieve high-quality model training. However, this distributed nature also makes FL highly …

Federated Learning

TrojanZoo: Towards Unified, Holistic, and Practical Evaluation of Neural Backdoors

2020-12-16 · Ren Pang, Zheng Zhang, Xiangshan Gao, Zhaohan Xi 외

Neural backdoors represent one primary threat to the security of deep learning systems. The intensive research has produced a plethora of backdoor attacks/defenses, resulting in a constant arms race. However, due to the …

Fighting Gradients with Gradients: Dynamic Defenses against Adversarial Attacks

2021-05-18 · NeurIPS 2021 12 · Dequan Wang, An Ju, Evan Shelhamer, David Wagner 외

Adversarial attacks optimize against models to defeat defenses. Existing defenses are static, and stay the same once trained, even while attacks change. We argue that models should fight back, and optimize their defenses…