AI Paper Atlas

Research map / Label-efficient and robust learning / Semi-supervised learning

The Perils of Learning From Unlabeled Data: Backdoor Attacks on Semi-supervised Learning

Virat Shejwalkar, Lingjuan Lyu, Amir Houmansadr

ICCV 2023 · Label-efficient and robust learning › Semi-supervised learning · approach: Safe and realistic semi-supervised learning · 12 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 clean-label static full-image repetitive trigger instead of DeHiB/Narcissus/CL-BadnetsSSL backdoor poisoning (FixMatch) · CIFAR10 (0.2% poison)ASR (%): 1.1 (DeHiB) -> 88.1Improved
use dirty-label adversarial-perturbation poisoning (DeHiB)SSL backdoor attack · CIFAR10ASR (%): 0.9 - 1.2 (DeHiB without D^l)Hurt
use small trigger patch (CL-Badnets) instead of full-image triggerSSL backdoor attack · CIFAR100 / CIFAR10ASR (%): ASR stays below 35%Hurt
apply strong augmentation to supervised training (ablation)CL-Badnets · CIFAR10 supervisedASR (%): 83.4 -> 0.0Hurt
apply strong augmentation to supervised training (ablation)Our attack · CIFAR10 supervisedASR (%): 99.8 -> 88.9Hurt
apply strong augmentation to supervised training (ablation)Narcissus · CIFAR10 supervisedASR (%): 100.0 -> 99.5No clear effect
add the backdoor trigger to target-class unlabeled dataFixMatch · CIFAR10clean accuracy (%): 91.1 -> 93.5Improved
vary poison size outside 0.2-0.4% of training data (ablation)Our attack · CIFAR10/SVHN ablationASR (%): ASR > 60% only within 0.2-0.4%Hurt
increase labeled data size (ablation)Our attack · CIFAR10 / SVHNASR (%): ASRs remain above 70%Mixed
combine Narcissus trigger with static trigger (ablation)Our attack · SSL backdoorASRHurt
duplicate poisoning data K times (ablation)Our attack · CIFAR10 UDA / FlexMatchASR (%): 81.5 -> 84.3 (UDA), 87.9 -> 89.1 (FlexMatch)Mixed
use interpolation-based poisoning (ablation)Our attack · CIFAR10ASRHurt

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