AI Paper Atlas

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

Red PANDA: Disambiguating Image Anomaly Detection by Removing Nuisance Factors

Niv Cohen, Jonathan Kahana, Yedid Hoshen

ICLR 2023 · Label-efficient and robust learning › OOD and anomaly detection · approach: Unified multi-class anomaly 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
Use nuisance-domain-guided representation disentanglement instead of DN2Image anomaly detection · Cars3DAD-Score (ROC-AUC): 0.946 -> 0.985Improved
Use nuisance-domain-guided representation disentanglement instead of DN2Image anomaly detection · Cars3DPA-Score (ROC-AUC; lower is better): 0.564 -> 0.506Improved
Use nuisance-domain-guided representation disentanglement instead of MeanShiftedImage anomaly detection · Cars3DRA-Score (ROC-AUC): 0.917 -> 0.980Improved
Use nuisance-domain-guided representation disentanglement instead of MeanShiftedImage anomaly detection · SmallNORBAD-Score (ROC-AUC): 0.948 -> 0.953Improved
Use nuisance-domain-guided representation disentanglement instead of CSIImage anomaly detection · SmallNORBPA-Score (ROC-AUC; lower is better): 0.556 -> 0.581Hurt
Use nuisance-domain-guided representation disentanglement instead of MeanShiftedImage anomaly detection · SmallNORBRA-Score (ROC-AUC): 0.811 -> 0.943Improved
Use nuisance-domain-guided representation disentanglement instead of DN2Image anomaly detection · Edges2ShoesAD-Score (ROC-AUC): 0.500 -> 0.781Improved
Use nuisance-domain-guided representation disentanglement instead of DN2Image anomaly detection · Edges2ShoesPA-Score (ROC-AUC; lower is better): 0.631 -> 0.711Hurt
Use nuisance-domain-guided representation disentanglement instead of SimCLRImage anomaly detection · Edges2ShoesRA-Score (ROC-AUC): 0.510 -> 0.719Improved
Replace separate per-domain contrastive losses with a single contrastive loss (ablation)Red PANDA · Cars3DPA-Score (ROC-AUC; lower is better): 0.506 -> 0.519Hurt
Replace separate per-domain contrastive losses with a single contrastive loss (ablation)Red PANDA · Edges2ShoesRA-Score (ROC-AUC): 0.719 -> 0.510Hurt

Later papers that compared with this method

1 later accepted paper reports Red PANDA as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

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