홈 › Papers

Adversarial Training for Free!

2019-04-29 · NeurIPS 2019 12 · Ali Shafahi, Mahyar Najibi, Amin Ghiasi, Zheng Xu, John Dickerson, Christoph Studer, Larry S. Davis, Gavin Taylor, Tom Goldstein

Adversarial training, in which a network is trained on adversarial examples, is one of the few defenses against adversarial attacks that withstands strong attacks. Unfortunately, the high cost of generating strong adversarial examples makes standard adversarial training impractical on large-scale problems like ImageNet. We present an algorithm that eliminates the overhead cost of generating adversarial examples by recycling the gradient information computed when updating model parameters. Our "free" adversarial training algorithm achieves comparable robustness to PGD adversarial training on the CIFAR-10 and CIFAR-100 datasets at negligible additional cost compared to natural training, and can be 7 to 30 times faster than other strong adversarial training methods. Using a single workstation with 4 P100 GPUs and 2 days of runtime, we can train a robust model for the large-scale ImageNet classification task that maintains 40% accuracy against PGD attacks. The code is available at https://github.com/ashafahi/free_adv_train.

📄 PDF Abstract BibTeX arXiv:1904.12843

Code (6)

ashafahi/free_adv_train 공식 구현 tf

mahyarnajibi/FreeAdversarialTraining 공식 구현 pytorch

AlanChou/Adversarial-Training-for-Free pytorch

a1600012888/YOPO-You-Only-Propagate-Once pytorch

locuslab/fast_adversarial pytorch

simon0987/Fast_FGSM pytorch

Tasks

Adversarial Attack Adversarial Defense Domain Generalization General Classification

Stability and Generalization in Free Adversarial Training

2024-04-13 · Xiwei Cheng, Kexin Fu, Farzan Farnia

While adversarial training methods have significantly improved the robustness of deep neural networks against norm-bounded adversarial perturbations, the generalization gap between their performance on training and test …

Generalization Bounds

DataFreeShield: Defending Adversarial Attacks without Training Data

2024-06-21 · Hyeyoon Lee, Kanghyun Choi, Dain Kwon, Sunjong Park 외

Recent advances in adversarial robustness rely on an abundant set of training data, where using external or additional datasets has become a common setting. However, in real life, the training data is often kept private …

Adversarial Robustness Dataset Generation

DAD++: Improved Data-free Test Time Adversarial Defense

2023-09-10 · Gaurav Kumar Nayak, Inder Khatri, Shubham Randive, Ruchit Rawal 외

With the increasing deployment of deep neural networks in safety-critical applications such as self-driving cars, medical imaging, anomaly detection, etc., adversarial robustness has become a crucial concern in the relia…

Adversarial Defense Adversarial Robustness Anomaly Detection Data-free Knowledge Distillation+4

A3: Active Adversarial Alignment for Source-Free Domain Adaptation

2024-09-27 · Chrisantus Eze, Christopher Crick

Unsupervised domain adaptation (UDA) aims to transfer knowledge from a labeled source domain to an unlabeled target domain. Recent works have focused on source-free UDA, where only target data is available. This is chall…

Active Learning Domain Adaptation Self-Supervised Learning Source-Free Domain Adaptation+1

FreeLB: Enhanced Adversarial Training for Natural Language Understanding

2019-09-25 · ICLR 2020 1 · Chen Zhu, Yu Cheng, Zhe Gan, Siqi Sun 외

Adversarial training, which minimizes the maximal risk for label-preserving input perturbations, has proved to be effective for improving the generalization of language models. In this work, we propose a novel adversaria…

ARC Natural Language Understanding Overall - Test Word Embeddings