AI Paper Atlas

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

Detecting and Diagnosing Adversarial Images with Class-Conditional Capsule Reconstructions

Yao Qin, Nicholas Frosst, Sara Sabour, Colin Raffel, Garrison Cottrell, Geoffrey Hinton

ICLR 2020 · Adversarial robustness and security › Attacks on text and graphs · approach: Biologically inspired robust vision · 54 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 CapsNet class-conditional reconstructions instead of CNN+CRAdversarial-image detection · MNIST, white-box targeted PGDundetected rate: 0 -> 0No clear effect
Use CapsNet rather than CNN+CRAdversarial-image detection · Fashion-MNIST and SVHN, targeted and untargeted R-PGDundetected rate: 84.4 -> 29.8; 86.0 -> 75.5; 97.9 -> 79.2; 99.5 -> 97.5Improved
Add class-conditional masking to CNN reconstruction (CNN+CR instead of CNN+R) (ablation)CNN adversarial-image detector · White-box targeted R-PGDundetected rate: 71.2 -> 68.1; 88.4 -> 84.4; 99.3 -> 97.9Improved
Remove class-conditional masking and reconstruct using all capsules (CapsNet All) (ablation)CapsNet detector · CIFAR-10, white-box R-PGDundetected rateHurt
Replace capsule masking with the winning-capsule vector as reconstruction input (DeepCaps) (ablation)CapsNet detector · CIFAR-10, white-box R-PGDundetected rateHurt
Use two-stage optimization instead of a single combined loss for targeted reconstructive attacks (ablation)Reconstructive attack · Targeted reconstructive attacksattack success rateImproved
Add reconstruction-loss minimization to a standard PGD attack (R-PGD)White-box adversarial attack · MNIST, Fashion-MNIST, and SVHNattack success rateHurt

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