paper-with-me

홈 › Papers

Scaling Adversarial Training to Large Perturbation Bounds

2022-10-18 · Sravanti Addepalli, Samyak Jain, Gaurang Sriramanan, R. Venkatesh Babu

The vulnerability of Deep Neural Networks to Adversarial Attacks has fuelled research towards building robust models. While most Adversarial Training algorithms aim at defending attacks constrained within low magnitude Lp norm bounds, real-world adversaries are not limited by such constraints. In this work, we aim to achieve adversarial robustness within larger bounds, against perturbations that may be perceptible, but do not change human (or Oracle) prediction. The presence of images that flip Oracle predictions and those that do not makes this a challenging setting for adversarial robustness. We discuss the ideal goals of an adversarial defense algorithm beyond perceptual limits, and further highlight the shortcomings of naively extending existing training algorithms to higher perturbation bounds. In order to overcome these shortcomings, we propose a novel defense, Oracle-Aligned Adversarial Training (OA-AT), to align the predictions of the network with that of an Oracle during adversarial training. The proposed approach achieves state-of-the-art performance at large epsilon bounds (such as an L-inf bound of 16/255 on CIFAR-10) while outperforming existing defenses (AWP, TRADES, PGD-AT) at standard bounds (8/255) as well.

📄 PDF Abstract BibTeX arXiv:2210.09852

Code (1)

val-iisc/oaat 공식 구현 pytorch

Tasks

Adversarial DefenseAdversarial Robustness

Methods 이 논문이 사용한 방법론

FLIP https://developer.nvidia.com/blog/flip-a-difference-evaluator-for-alternating-images/
ALIGN In the ALIGN method, visual and language representations are jointly trained from noisy image alt-text data. The image and text encoders are learned via contrastive loss…

Similar Papers 제목 키워드 기반

Towards Achieving Adversarial Robustness Beyond Perceptual Limits

2021-09-29 · ICML Workshop AML 2021 7 · Sravanti Addepalli, Samyak Jain, Gaurang Sriramanan, Venkatesh Babu Radhakrishnan

The vulnerability of Deep Neural Networks to Adversarial Attacks has fuelled research towards building robust models. While most Adversarial Training algorithms aim towards defending attacks constrained within low magnit…

Adversarial DefenseAdversarial Robustness

Scaling provable adversarial defenses

2018-05-31 · NeurIPS 2018 12 · Eric Wong, Frank R. Schmidt, Jan Hendrik Metzen, J. Zico Kolter

Recent work has developed methods for learning deep network classifiers that are provably robust to norm-bounded adversarial perturbation; however, these methods are currently only possible for relatively small feedforwa…

Scaling up the randomized gradient-free adversarial attack reveals overestimation of robustness using established attacks

2019-03-27 · Francesco Croce, Jonas Rauber, Matthias Hein

Modern neural networks are highly non-robust against adversarial manipulation. A significant amount of work has been invested in techniques to compute lower bounds on robustness through formal guarantees and to build pro…

Adversarial Attack

Adaptive Smoothness-weighted Adversarial Training for Multiple Perturbations with Its Stability Analysis

2022-10-02 · Jiancong Xiao, Zeyu Qin, Yanbo Fan, Baoyuan Wu 외

Adversarial Training (AT) has been demonstrated as one of the most effective methods against adversarial examples. While most existing works focus on AT with a single type of perturbation e.g., the $\ell_\infty$ attacks)…

Adversarial Robustness

Certified Adversarial Robustness Within Multiple Perturbation Bounds

2023-04-20 · Soumalya Nandi, Sravanti Addepalli, Harsh Rangwani, R. Venkatesh Babu

Randomized smoothing (RS) is a well known certified defense against adversarial attacks, which creates a smoothed classifier by predicting the most likely class under random noise perturbations of inputs during inference…

Adversarial Robustness