AI Paper Atlas

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

Enhancing Adversarial Contrastive Learning via Adversarial Invariant Regularization

Xilie Xu, Jingfeng Zhang, Feng Liu, Masashi Sugiyama, Mohan Kankanhalli

NeurIPS 2023 · Adversarial robustness and security › Certified and empirical robustness · approach: Robust overfitting and regularization · 4 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 AIR regularization to ACLACL · CIFAR-10 SLFstandard accuracy SA (%): 78.27 -> 80.03Improved
add AIR regularization to DynACLDynACL · CIFAR-10 ALFstandard accuracy SA (%): 72.90 -> 77.42Improved
add AIR regularization to DynACLDynACL · CIFAR-100 ALFstandard accuracy SA (%): 43.58 -> 45.63Improved
add AIR regularization to DynACLDynACL · STL-10 SLFrobust accuracy AA (%): 46.49 -> 47.66Improved
add AIR to ACL, cross-task transferACL · CIFAR-10 -> CIFAR-100 ALFstandard accuracy SA (%): 28.58 -> 36.10Improved
add AIR to ACL, cross-task transferACL · CIFAR-10 -> CIFAR-100 ALFrobust accuracy AA (%): 11.09 -> 13.30Improved
add AIR to ACL, common corruptionsACL · CIFAR-10-C SLF CS-1accuracy under common corruption (%): 76.57 -> 78.55Improved
add AIR to ACL, cross-task corruptionsACL · CIFAR-10 -> CIFAR-100-C SLF CS-1accuracy under common corruption (%): 31.39 -> 37.82Improved
add AIR to DynACL after LP-AFF post-processingDynACL++ · CIFAR-10 SLFstandard accuracy SA (%): 79.81 -> 81.80Improved
add AIR to DynACL on high-resolution ImagenetteDynACL · Imagenette SLF ResNet-50standard accuracy SA (%): 80.74 -> 81.66Improved
add adversarial AIR term only (lambda1=0, lambda2=0.5) (ablation)ACL-AIR · CIFAR-10 SLFrobust accuracy AA (%): 37.39 -> 38.70Improved
add SIR term only (lambda1=0.5, lambda2=0) (ablation)ACL-AIR · CIFAR-10 SLFrobust accuracy AA (%): 37.39 -> 37.51No clear effect

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