Research map / Label-efficient and robust learning / OOD and anomaly detection
Explicit Boundary Guided Semi-Push-Pull Contrastive Learning for Supervised Anomaly Detection
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| Use BGAD instead of NFAD | MVTecAD multi-class anomaly detection · MVTecAD multi-class | image-level AUROC: 0.968 -> 0.993 | Helped |
| Use BGAD instead of NFAD | MVTecAD multi-class anomaly localization · MVTecAD multi-class | PRO: 0.946 -> 0.976 | Helped |
| Use BGAD instead of NFAD | BTAD anomaly localization · BTAD | pixel-level AUROC: 0.978 -> 0.986 | Helped |
| Use BGAD instead of NFAD | MVTecAD one-class generalization · MVTecAD one-class | image-level AUROC: 0.998 -> 0.999 | Helped |
| Replace BG-SPP with conventional contrastive loss (ablation) | BGAD · MVTecAD one-class | image-level AUROC: 0.999 -> 0.868 | Hurt |
| Add RPAG pseudo-anomaly generation (ablation) | BGAD with 5 training anomalies · MVTecAD | image AUROC: 0.982 -> 0.985 | Helped |
| Add RPAG pseudo-anomaly generation (ablation) | BGAD with 10 training anomalies · MVTecAD | image AUROC: 0.983 -> 0.988 | Helped |
| Add asymmetric weighting (ablation) | BGAD with RPAG and 10 training anomalies · MVTecAD | image AUROC: 0.988 -> 0.993 | Helped |
| Use RPAG instead of DRAEM synthetic anomaly generation (ablation) | BGAD · MVTecAD | image AUROC: 0.970 -> 0.988 | Helped |
| Increase the BG-SPP loss weight from 1 to 10 (ablation) | BGAD · MVTecAD hard subsets | PRO: 0.945 -> 0.950 | Helped |
| Train on easy subsets and test on hard unseen anomalies (ablation) | BGAD versus NFAD · MVTecAD unseen subsets | PRO: 0.863 -> 0.930 | Helped |
Closest papers by meaning
- Generating and Reweighting Dense Contrastive Patterns for Unsupervised Anomaly Detection (AAAI 2024 · similarity 0.87)
- Normal-Abnormal Guided Generalist Anomaly Detection (NeurIPS 2025 · similarity 0.87)
- One-to-Normal: Anomaly Personalization for Few-shot Anomaly Detection (NeurIPS 2024 · similarity 0.86)
- Beyond Normal References: Discriminative Few-Shot Anomaly Detection (NeurIPS 2026 · similarity 0.86)
- ResAD: A Simple Framework for Class Generalizable Anomaly Detection (NeurIPS 2024 · similarity 0.86)
- ADPretrain: Advancing Industrial Anomaly Detection via Anomaly Representation Pretraining (NeurIPS 2025 · similarity 0.86)
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.