AI Paper Atlas

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

Dual-branch Robust Unlearnable Examples

Xianlong Wang, Hangtao Zhang, Wenbo Pan, Ziqi Zhou, Changsong Jiang, Li Zeng, Xiaohua Jia

ICML 2026 · Adversarial robustness and security › Certified and empirical robustness · approach: Adversarial training recipes · 0 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 unlearnability-enhancing ensemble (UEE) to spatial+color branches (ablation)DUNE · CIFAR-10 ResNet18average test accuracy over defenses (%): 31.15 -> 19.67Improved
add UEE to spatial+color branches (ablation)DUNE · CIFAR-10 VGG19average test accuracy over defenses (%): 27.11 -> 18.09Improved
add color branch to spatial branch plus ensemble (ablation)DUNE · CIFAR-10 ResNet18average test accuracy over defenses (%): 50.20 -> 19.67Improved
add spatial branch to color branch plus ensemble (ablation)DUNE · CIFAR-10 ResNet18average test accuracy over defenses (%): 32.68 -> 19.67Improved
add color and spatial branches and ensemble to spatial-only (ablation)DUNE · CIFAR-10 ResNet18average test accuracy over defenses (%): 57.05 -> 19.67Improved
add spatial branch and ensemble to color-only (ablation)DUNE · CIFAR-10 ResNet18average test accuracy over defenses (%): 43.67 -> 19.67Improved
add color branch to spatial branch under ISS-J (ablation)DUNE (vs spatial + Region-4) · CIFAR-10 ResNet18test accuracy under ISS-J (%): 80.97 -> 28.88Improved
add DUNE instead of best baseline CUDA under no-defense/ISS-J averageDUNE · CIFAR-10 ResNet18average test accuracy (%): 39.58 -> 23.29Improved
use PSO color-branch optimization instead of PGDDUNE · CIFAR-10total generation time (hrs): 0.3459 -> 0.9574Hurt
use PSO color-branch optimization instead of PGDDUNE · ImageNet-100total generation time (hrs): 1.3641 -> 7.5766Hurt
increase beta excessively (ablation)DUNE spatial branch · CIFAR-10 ResNet18test accuracyHurt
apply DUNE under ISS-J instead of EMDUNE · CIFAR-10 ResNet18test accuracy under ISS-J (%): 78.91 -> 28.88Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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