AI Paper Atlas

Research map / Adversarial robustness and security / Black-box and transfer attacks

Defending against Data-Free Model Extraction by Distributionally Robust Defensive Training

Zhenyi Wang, Li Shen, Tongliang Liu, Tiehang Duan, Yanjun Zhu, Donglin Zhan, David Doermann, Mingchen Gao

NeurIPS 2023 · Adversarial robustness and security › Black-box and transfer attacks · approach: Data-free substitute training

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
apply MeCo defense instead of no defenseDFME attack · CIFAR10, clone ResNet18-8Xclone accuracy: 87.36 ± 0.78% -> 51.68 ± 1.96%Improved
apply MeCo instead of GRAD baselineDFME attack · CIFAR100, clone ResNet18-8Xclone accuracy: 39.07 ± 2.72% -> 29.57 ± 1.97%Improved
apply MeCo instead of no defenseMAZE attack · CIFAR10, clone ResNet18-8Xclone accuracy: 45.17 ± 0.73% -> 21.89 ± 2.07%Improved
apply MeCo instead of no defenseDFMS-HL decision-based attack · CIFAR100, clone ResNet18-8Xclone accuracy: 72.57 ± 1.28% -> 59.30 ± 1.70%Improved
output-perturbation baselines (RandP, P-poison, GRAD) instead of no defenseDFMS-HL decision-based attack · CIFAR10, clone ResNet18-8Xclone accuracy: 84.67 ± 1.90% -> 84.28 ± 1.95%No clear effect
replace MeCo's small l1 perturbation with fixed l1 budget 1.0 (GRAD)target model · CIFAR10 utilityl1 norm of probability perturbation: 1.0 -> 0.099Improved
MeCo input perturbation on benign accuracytarget model · CIFAR10test accuracy: 94.91 ± 0.37% -> 94.17 ± 0.56%Hurt
attacker adapts by adding its own random perturbation generatorMeCo vs DFME · CIFAR10, clone ResNet18-8Xclone accuracy: 51.68 ± 1.96% -> 25.79 ± 0.81%Improved
score-based attacker uses only hard label instead of probabilitiesMeCo vs DFME · CIFAR10, clone ResNet18-8Xclone accuracy: 51.68 ± 1.96% -> 41.23 ± 0.58%Improved
add DRO to defensive training (ablation)MeCo · target model utilitymodel utility: 5.3% - 5.6%Improved
increase gamma (ablation)MeCo · CIFAR10utility vs defenseMixed
increase number of test-data simulation steps Q (ablation)MeCo · benign accuracybenign accuracyHurt

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