AI Paper Atlas

Research map / Label-efficient and robust learning / OOD and anomaly detection

RankOOD - Class Ranking-based Out-of-Distribution Detection

Dishanika Denipitiyage, Naveen Karunanayake, Suranga Seneviratne, Sanjay Chawla

CVPR 2026 · Label-efficient and robust learning › OOD and anomaly detection · approach: Spurious-aware OOD detection · 1 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
train with RankOOD-T (ListMLE ranking loss) and score with RankOOD-S instead of CE-trained MSPRankOOD · CIFAR-10 (ID) vs SVHN (OOD)FPR95 (%): 24.24 -> 12.23 (37.12% reduction)Improved
RankOOD vs strongest baseline on near-OOD ImageNet-200RankOOD · Near-OOD, ImageNet-200 as IDFPR95 (%): reduces FPR95 by 4.3% and improves AUROC by 0.50%Improved
RankOOD vs prior rank-based methods CRAFT and ExCeLRankOOD · Far-OOD and near-OOD averagesFPR95 (%): 7.51% far-OOD and 4.21% near-OOD average FPR95 reductionImproved
RankOOD vs GEN among methods without outliersRankOOD · Near-OOD, CIFAR-100 as IDFPR95 (%): outperforms GEN by 3.36%Improved
RankOOD vs G-ODIN on far-OOD TinyImageNetRankOOD · Far-OOD, ImageNet-200 as IDFPR95 (%): outperforms G-ODIN by 8.12%Improved
RankOOD vs OE (trained with outliers) on near-OOD averageRankOOD · Near-OOD averageFPR95 (%): 44.79 vs 34.29 (OE better)Hurt
train on top ranks only (vs adding bottom ranks) (ablation)RankOOD-T · CIFAR-100ID accuracyHurt
add bottom ranks to top ranks (Top-Bottom, top 10 and bottom 10, N=20) (ablation)RankOOD-T · CIFAR-100AUROC and ID accuracyImproved
use more ranks in Top-Bottom subset beyond N=20 (ablation)RankOOD-T · CIFAR-100AUROCHurt
sweep loss weight alpha in CE + alpha*ListMLE (ablation)RankOOD-T · CIFAR-10 and CIFAR-100AUROC: best alpha 0.8 for CIFAR-10 and 1.0 for CIFAR-100Mixed
raise reference logit percentile threshold Ref past 0.95 (ablation)RankOOD-S · CIFAR-10 and CIFAR-100AUROC: AUROC peaks around 0.95 then declinesMixed
train with RankOOD-T instead of standard CE under various logit-based OOD scores (ablation)RankOOD-T · OpenOOD benchmarkAUROCImproved

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