AI Paper Atlas

Research map / Adversarial robustness and security / Attacks on text and graphs

Sample Efficient Detection and Classification of Adversarial Attacks via Self-Supervised Embeddings

Mazda Moayeri, Soheil Feizi

ICCV 2021 · Adversarial robustness and security › Attacks on text and graphs · approach: Defenses against universal perturbations · 24 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 SimCLR self-supervised embeddings instead of supervised ResNet50 embeddingsSimCat · ImageNet 8-attack classification (5 samples per attack)accuracy (%): +7.7Improved
use SimCLR embeddings instead of supervised embeddingsSimCat · ImageNet detection (5 samples per attack)accuracy (%): 68.5 (+6.4)Improved
use SimCLR embeddings instead of supervised embeddingsSimCat · SVHN PGD-l2 vs PGD-linf classification (25 samples)accuracy (%): 77.1 (+21.0)Improved
increase training samples per attack from 5 to 100SimCat · ImageNet 8-attack classificationaccuracy (%): 27.1 -> 58.1Improved
increase training samples per attack from 2 to 50SimCat · SVHN PGD-linf detectionaccuracy (%): 77.3 -> 94.2Improved
train on union of attacks with 5 samples each instead of one attack with 100 samplesSimCat · ImageNet unseen-attack detectionaccuracy (%): 64.7 -> 66.5Improved
train on union of attacks with 20 samples each instead of 5SimCat · ImageNet unseen-attack detectionaccuracy (%): 66.5 -> 71.1Improved
use ensemble of separate BP and CP detectors instead of one joint detector (ablation)SimCat · STL10 poison defense (avg PSR)poison success rate (%): 13.9 -> 9.7Improved
add SimCat poison defenseSTL10 linear transfer model · STL10 poison defense (avg PSR)poison success rate (%): 21.8 -> 9.7Improved
add SimCat ensemble poison defenseSTL10 linear transfer model · STL10 clean accuracyaccuracy (%): 86 -> 85Hurt
exclude FC poisons from general detection (ablation)SimCat · STL10 general poison detectionaccuracy (%): 64.5 -> 70.5Improved
add adaptive adversarial training (ablation)SimCat · ImageNet detection under adaptive PGD-l2robustness (%): 39.25 -> 67.95Improved

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