AI Paper Atlas

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

Adversarially Robust Distillation

Micah Goldblum, Liam Fowl, Soheil Feizi, Tom Goldstein

AAAI 2020 · Adversarial robustness and security › Certified and empirical robustness · approach: Adversarial training with feature consistency · 227 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
Use ARD instead of clean-image distillation from a robust ResNet18 teacherMobileNetV2 · CIFAR-10robust accuracy (20-step PGD): 38.21% -> 50.22%Improved
Distill from an adversarially trained teacher using clean imagesResNet18 and MobileNetV2 students · CIFAR-10robust accuracy (20-step PGD): 0.0% -> 40.13%; 0.0% -> 38.21%Improved
Use clean-image distillation from a robust WRN teacher rather than ARDMobileNetV2 · CIFAR-10robust accuracy (20-step PGD): 5.46%; 21.69%Hurt
Use ARD from a TRADES WRN teacher instead of adversarial trainingMobileNetV2 · CIFAR-10robust accuracy (20-step PGD): 46.90% -> 50.42%Improved
Set α=0.95 instead of α=1 (ablation)TRADES WRN ARD → MobileNetV2 · CIFAR-10natural accuracy: 82.63% -> 84.70%Improved
Set α=0.95 instead of α=1 (ablation)TRADES WRN ARD → MobileNetV2 · CIFAR-10robust accuracy (20-step PGD): 50.42% -> 46.28%Hurt
Use ARD from a robust WRN teacher instead of adversarial trainingMobileNetV2 · CIFAR-100robust accuracy (20-step PGD): 22.80% -> 27.64%Improved
Use ARD distillation from ResNet18 instead of adversarial trainingResNet18 · CIFAR-10robust accuracy (20-step PGD): 44.46% -> 51.21%Improved
Use ARD distillation from MobileNetV2 instead of adversarial trainingMobileNetV2 · CIFAR-10robust accuracy (20-step PGD): 46.90% -> 47.95%Improved
Use Fast-ARD rather than free adversarial trainingMobileNetV2 (m=4) · CIFAR-10robust accuracy (20-step PGD): 23.13% -> 37.07%Improved
Replace full ARD with Fast-ARDMobileNetV2 · CIFAR-10training time: 41.06 -> 5.15 hrsImproved
Replace adversarial training with ARDMobileNetV2 · CIFAR-10 under MI-FGSM20robust validation accuracy: 50.82% -> 55.16%Improved

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