AI Paper Atlas

Research map / Adversarial robustness and security / Backdoor and poisoning

Randomized Channel Shuffling: Minimal-Overhead Backdoor Attack Detection without Clean Datasets

Ruisi Cai, Zhenyu Zhang, Tianlong Chen, Xiaohan Chen, Zhangyang Wang

NeurIPS 2022 · Adversarial robustness and security › Backdoor and poisoning · approach: Training-time backdoor defense · 3 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
Detect with randomized channel shuffling instead of Neural Cleansebackdoor detection · BadNets CIFAR-10 ResNet-20AUROC: 0.669 -> 1.000Improved
Detect with randomized channel shuffling instead of STRIPbackdoor detection · BadNets CIFAR-10 ResNet-20AUROC: 0.459 -> 1.000Improved
Detect with channel shuffling instead of Neural Cleansebackdoor detection · WaNet CIFAR-10 ResNet-18AUROC: 0.314 -> 0.954Improved
Detect with channel shuffling instead of STRIPbackdoor detection · BadNets GTSRB ResNet-20AUROC: 0.587 -> 0.880Improved
Move trigger to bottom left (ablation)Ours · BadNets CIFAR-10 ResNet-20AUROC: 1.000No clear effect
Random trigger location (ablation)Ours · BadNets CIFAR-10 ResNet-20AUROC: 1.000No clear effect
Use RGB trigger instead of gray-scale (ablation)Ours · BadNets CIFAR-10 ResNet-20AUROC: 1.000No clear effect
Increase number of target classes from 2 to 5 (ablation)Ours · BadNets CIFAR-10 ResNet-20AUROC: 0.978 -> 0.770Hurt
Source-label-specific (partial) backdoor (ablation)Ours · BadNets CIFAR-10 ResNet-20AUROC: 0.945Mixed
Retrain with randomized channel shufflingbackdoored ResNet-20 · BadNets CIFAR-10test accuracy: 91.4 ± 0.5% -> 16.2 ± 7.3%Hurt
Retrain with randomized channel shufflingbackdoored ResNet-20 · BadNets CIFAR-10attack success rate: 100.0 ± 0.0% -> 100.0 ± 0.0%No clear effect
Retrain with randomized channel shufflingbackdoored AlexNet · BadNets CIFAR-10test accuracy: 85.7 ± 0.5% -> 45.4 ± 6.7%Hurt

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