AI Paper Atlas

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

ReContrast: Domain-Specific Anomaly Detection via Contrastive Reconstruction

Jia Guo, shuai lu, LIze JIa, Weihang Zhang, Huiqi Li

NeurIPS 2023 · Label-efficient and robust learning › OOD and anomaly detection · approach: Unified multi-class anomaly detection · 27 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 global cosine distance instead of regional cosine distance (ablation)RD4AD-style reconstruction · MVTec ADI-AUROC (%): 95.31 -> 98.86Improved
jointly optimize encoder and decoder without safeguards (ablation)feature reconstruction · MVTec ADI-AUROC (%): 98.86 -> 91.54Hurt
add stop-gradient when optimizing encoder (ablation)Config.C · MVTec ADI-AUROC (%): 91.54 -> 94.64Improved
add contrastive pairs (frozen plus trained encoder views) (ablation)Config.C · MVTec ADI-AUROC (%): 91.54 -> 97.59Improved
combine stop-gradient and contrastive pairs (ablation)Config.C · MVTec ADI-AUROC (%): 91.54 -> 99.13Improved
add hard-normal mining (L_global-hm) (ablation)Config.E · MVTec ADI-AUROC (%): 99.13 -> 99.45Improved
add hard-normal mining to frozen baseline (ablation)Config.B · MVTec ADI-AUROC (%): 98.86 -> 99.00Improved
remove global cosine loss from full pipeline (ablation)Config.E · APTOSI-AUROC (%): 96.39 -> 97.32Improved
add global cosine loss to Config.E (ablation)Config.E · MVTec ADI-AUROC (%): 98.93 -> 99.13Improved
raise hard-mining threshold alpha from 1 to 2 (ablation)ReContrast · MVTec ADI-AUROC (%): 99.45 -> 98.85Hurt
use two encoders at inference (frozen plus trained) instead of trained only (ablation)ReContrast · MVTec ADI-AUROC (%): 99.38 -> 99.45No clear effect
use larger backbone WideResNet50 versus ResNet18 (ablation)ReContrast · MVTec ADI-AUROC (%): 98.7 -> 99.5Improved

Later papers that compared with this method

6 later accepted papers report ReContrast as a baseline in their results tables; 4 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