AI Paper Atlas

Research map / Adversarial robustness and security / Certified and empirical robustness

Make Some Noise: Reliable and Efficient Single-Step Adversarial Training

Pau de Jorge, Adel Bibi, Riccardo Volpi, Amartya Sanyal, Philip Torr, Grégory Rogez, Puneet K. Dokania

NeurIPS 2022 · Adversarial robustness and security › Certified and empirical robustness · approach: Adversarial training refinements · 31 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add strong noise augmentation and remove clipping around clean sample (N-FGSM vs RS-FGSM)FGSM adversarial training · ImageNet eps=6/255PGD50-10 accuracy (%): 16.49 -> 17.12Improved
add noise augmentation (N-FGSM) to FGSMFGSM · ImageNet eps=6/255PGD50-10 accuracy (%): 0.08 -> 17.12Improved
clip perturbation around the clean sample (ablation)N-FGSM · CIFAR-10 eps=8/255PGD50-10 accuracyHurt
increase noise magnitude k beyond eps (ablation)N-FGSM (no clipping) · CIFAR-10 large epsavoids catastrophic overfittingImproved
increase step size alpha from 1.25eps to 1.5eps keeping k=1eps (ablation)N-FGSM · CIFAR-10 eps=8/255PGD50-10 accuracy (%): 49.68% -> 0Hurt
add noise to FGSM step (vs larger alpha without noise) (ablation)FGSM · CIFAR-10 eps=8/255PGD50-10 accuracyImproved
increase training eps from 8/255 to 12/255 (ablation)RS-FGSM · CIFAR-10 eps=8/255 testPGD50-10 accuracy (%): 46.08 -> 0.0Hurt
increase training eps from 8/255 to 12/255 (ablation)GradAlign · CIFAR-10 eps=8/255 testPGD50-10 accuracy (%): 48.14 -> 50.6 robust; 81.9 -> 73.29 cleanMixed
use long training schedule with final model instead of fast schedule (ablation)N-FGSM · CIFAR-10 eps=8/255PGD50-10 accuracy (%): 48.12 -> 36.56Hurt
add early stopping (best model) to long schedule (ablation)N-FGSM · CIFAR-10 eps=8/255PGD50-10 accuracy (%): 36.56 -> 48.48Improved
use N-FGSM instead of GradAlignsingle-step AT · CIFAR-10/100, SVHN PreActResNet18robust accuracy and cost: 3x speed-upImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy