Research map / Detection and segmentation / Open-set and few-shot detection
Defect Cue-Preserved Structural Feature Refinement for Few-Shot Anomaly Detection
Le Jiang, Yan Huang, Zhen Xu, Yong Xu, Hau-San Wong, Si Wu
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add conditional anomaly cue amplification (CACA) (ablation) | DCP-SFR · VisA 1-shot | p-F1-max: 43.2 -> 44.6 | Improved |
| add CACA (ablation) | DCP-SFR · VisA 1-shot | i-AUROC: 91.5 -> 92.0 | Improved |
| add anomaly cue-guided localization (ACL) (ablation) | DCP-SFR · VisA 1-shot | p-AP: 37.7 -> 39.9 | Improved |
| add ACL (ablation) | DCP-SFR · VisA 1-shot | i-AUROC: 89.7 -> 92.0 | Improved |
| add structure-aware segmentation refinement (SSR) (ablation) | DCP-SFR · VisA 1-shot | p-AUROC: 97.1 -> 98.2 | Improved |
| add SSR (ablation) | DCP-SFR · VisA 1-shot | p-F1-max: 41.0 -> 44.6 | Improved |
| increase reference shots from 1 to 4 | DCP-SFR · VisA | p-F1-max: 44.9 -> 47.1 | Improved |
| use DCP-SFR instead of AnomalyDINO | few-shot anomaly detection · MVTec AD 1-shot | p-AP: 56.5 -> 61.2 | Improved |
| use DCP-SFR instead of AnomalyDINO | few-shot anomaly detection · VisA 1-shot | p-F1-max: 45.1 -> 44.9 | Hurt |
Closest papers by meaning
- Few Shot Part Segmentation Reveals Compositional Logic for Industrial Anomaly Detection (AAAI 2024 · similarity 0.87)
- 3CAD: A Large-Scale Real-World 3C Product Dataset for Unsupervised Anomaly Detection (AAAI 2025 · similarity 0.87)
- Exploring Intrinsic Normal Prototypes within a Single Image for Universal Anomaly Detection (CVPR 2025 · similarity 0.87)
- MVREC: A General Few-shot Defect Classification Model Using Multi-View Region-Context (AAAI 2025 · similarity 0.87)
- RealNet: A Feature Selection Network with Realistic Synthetic Anomaly for Anomaly Detection (CVPR 2024 · similarity 0.86)
- Towards Generalizable 3D Anomaly Detection via Relational Inconsistency Modeling (NeurIPS 2026 · 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.
