AI Paper Atlas

Research map / Adversarial robustness and security / Certified and empirical robustness

Sparse-PGD: A Unified Framework for Sparse Adversarial Perturbations Generation

Xuyang Zhong, Chen Liu

TPAMI 2025 · Adversarial robustness and security › Certified and empirical robustness · approach: Attack evaluation and defense critique · 10 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
add decomposition delta = p*m to PGD0 without restart (ablation)PGD0 · CIFAR-10 Fast-EG-l1robust accuracy: 49.4 -> 58.0Hurt
add continuous mask and l0 projection (ablation)sPGD_proj · CIFAR-10 Fast-EG-l1robust accuracy: 58.0 -> 33.9Improved
add random reinitialization (ablation)sPGD_proj · CIFAR-10 Fast-EG-l1robust accuracy: 33.9 -> 18.1Improved
vary step size alpha for magnitude from 1/16 to 1 (ablation)sPGD_proj · CIFAR-10 Fast-EG-l1robust accuracy: 19.6 -> 18.1 -> 19.0No clear effect
vary attack reinitialization tolerance t from 1 to 10 (ablation)sPGD_proj · CIFAR-10 Fast-EG-l1robust accuracy: 18.1 -> 18.5No clear effect
random backward function policy instead of always projected (ablation)sTRADES · CIFAR-10 sAArobust accuracy: 41.5 -> 61.7Improved
alternate backward function every 5 epochs instead of random (ablation)sTRADES · CIFAR-10 sAArobust accuracy: 61.7 -> 51.5Hurt
train with 6x larger epsilon than test (multi-epsilon) (ablation)sTRADES · CIFAR-10 sAArobust accuracy: 34.0 -> 61.7Improved
increase multi-epsilon from 6x to 10x (ablation)sTRADES · CIFAR-10 sAArobust accuracy: 61.7 -> 55.5Hurt
raise training reinitialization tolerance from 3 to 10 (ablation)sTRADES · CIFAR-10 sAArobust accuracy: 51.7 -> 61.7Improved
multi-epsilon instead of multi-size strategy for 3x3 patch training (ablation)sTRADES · CIFAR-10 3x3 patchsAA robust accuracy: 67.9 -> 72.3Improved
use sTRADES instead of sATsPGD adversarial training · CIFAR-10 eps=20sAA robust accuracy: 36.2 -> 61.7Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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