AI Paper Atlas

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

SPot-the-Difference Self-Supervised Pre-training for Anomaly Detection and Segmentation

Yang Zou, Jongheon Jeong, Latha Pemula, Dongqing Zhang, Onkar Dabeer

ECCV 2022 · Label-efficient and robust learning › OOD and anomaly detection · approach: Unified multi-class anomaly detection · 752 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 SPD regularizationSimSiam pre-training · VisA 1-class (PaDiM) classificationAU-PR: 80.2 -> 82.8Improved
add SPD regularizationSimSiam pre-training · VisA 2-class high-shot segmentationAU-PR: 53.8 -> 59.7Improved
add SPD regularizationsupervised pre-training · VisA 2-class high-shot segmentationAU-PR: 65.1 -> 71.9Improved
add SPD regularizationMoCo pre-training · VisA 1-class (PaDiM) classificationAU-PR: 83.6 -> 84.1Improved
add SPD regularizationSimCLR pre-training · VisA 2-class high-shot classificationAU-PR: 93.4 -> 92.7Hurt
add SPD regularizationSimCLR pre-training · MVTec-AD 1-class (PaDiM) classificationAU-PR: 94.7 -> 96.8Improved
add SPD regularizationSimSiam pre-training · VisA 2-class 5-shot classificationAU-PR: 51.9 -> 56.1Improved
add SPD regularizationMoCo pre-training · VisA 2-class 5-shot segmentationAU-ROC: 80.5 -> 78.5Hurt
use SPD loss weight 0.1 instead of 0.5 (ablation)SimSiam+SPD · VisA 1-class (PaDiM) classificationAU-PR: 80.5 -> 82.8Improved
use SmoothBlend instead of CutPaste (ablation)SimSiam+SPD · VisA 1-class (PaDiM) classificationAU-PR: 78.8 -> 82.8Improved
use SPD loss instead of cross-entropy loss (ablation)SimSiam+SPD · VisA 1-class (PaDiM) classificationAU-PR: 71.4 -> 82.8Improved
add SPD regularizationsupervised Wide ResNet50, PatchCore · VisA 1-class classificationAU-PR: 93.3 -> 93.8Improved

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