AI Paper Atlas

Research map / Adversarial robustness and security / Backdoor and poisoning

CatchBackdoor: Backdoor Detection via Critical Trojan Neural Path Fuzzing

Haibo Jin, Ruoxi Chen, Jinyin Chen, Haibin Zheng, Yang Zhang, Haohan Wang

ECCV 2024 · 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 trojans via fuzzed critical trojan path instead of trigger-size-dependent reverse engineeringCatchBackdoor · CIFAR-10 AlexNet, BadNets large triggersLCR: K-Arm N at 18x18 vs CatchBackdoor 72.8% at 18x18Improved
CatchBackdoor vs baselines on stealthy attacks (blending, defense adaptive)CatchBackdoor · MNIST LeNet-5, Poison frogsLCR: 84.4% -> 93.0%Improved
CatchBackdoor on clean-label attacks BP and Sleeper Agent versus K-ArmCatchBackdoor · MNIST LeNet-4 and CIFAR-10 AlexNetLCR: BP LeNet-4: 85.0% (K-Arm) vs 83.2%Hurt
increase model depth and dataset scaleCatchBackdoor · a-ImageNet VGG19LCR: LCR around 76% on averageHurt
start fuzzing from random noise instead of benign examples (ablation)CatchBackdoor · MNIST, CIFAR-10, a-ImageNetLCR: around 73% on averageNo clear effect
adaptive attacks S-AA and I-AA against CatchBackdoorCatchBackdoor · adaptive attacksLCR: LCR over 85%No clear effect
increase top-k neurons k1 and k2 (ablation)CatchBackdoor · VGG19 a-ImageNetLCRMixed
use single neuron (k=1) rather than multi-neuron path against ABE (ablation)CatchBackdoor · VGG19 a-ImageNet, ABELCRHurt
replace top-k neural path of original class with target class contributions (ablation)LeNet-5 on MNIST · MNIST target class 1consistency rate: 96.6%Improved
LCR threshold between 0.4 and 0.6 (ablation)CatchBackdoor · 100 benign and trojaned modelsAUC: AUC larger than 0.9Improved

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