Research map / Label-efficient and robust learning / OOD and anomaly detection
PIRN: Prototypical-based Intra-modal Reconstruction with Normality Communication for Multi-modal Anomaly Detection.
YITING LI, Xulei Yang, Jing Zhang, Sichao Tian, Jingyi Liao, Fayao Liu
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 Balanced Prototype Assignment (ablation) | PIRN · MVTec-3D-AD (10-shot) | image AUROC: 0.883 -> 0.922 | Improved |
| add Adaptive Prototype Refinement (ablation) | PIRN · MVTec-3D-AD (10-shot) | image AUROC: 0.916 -> 0.922 | Improved |
| add Multimodal Normality Communication (ablation) | PIRN · MVTec-3D-AD (10-shot) | image AUROC: 0.867 -> 0.922 | Improved |
| add surface normals to RGB (ablation) | PIRN · MVTec-3D-AD (10-shot) | image AUROC: 0.827 -> 0.922 | Improved |
| use surface normals instead of RGB alone (ablation) | PIRN · MVTec-3D-AD (10-shot) | image AUROC: 0.827 -> 0.879 | Improved |
| use balanced OT instead of global averaging for APR aggregation (ablation) | PIRN · MVTec-3D-AD (10-shot) | image AUROC: 0.915 -> 0.922 | Improved |
| use balanced OT instead of top-k averaging (ablation) | PIRN · MVTec-3D-AD (10-shot) | image AUROC: 0.921 -> 0.922 | No clear effect |
| increase prototypes from 10 to 100 (ablation) | PIRN · MVTec-3D-AD (all-shot) | image AUROC: 0.963 -> 0.901 | Hurt |
| reduce prototypes from 10 to 5 (ablation) | PIRN · MVTec-3D-AD (all-shot) | image AUROC: 0.963 -> 0.954 | Hurt |
| increase decoder layers from 2 to 8 (ablation) | PIRN · MVTec-3D-AD (10-shot) | image AUROC: 0.924 -> 0.869 | Hurt |
| use PIRN instead of INP-Former | INP-Former · MVTec-3D-AD (5-shot) | image AUROC: 0.851 -> 0.890 | Improved |
| use PIRN instead of FIND in compute | FIND · MVTec-3D-AD (10-shot) | FLOPs: 728.46 -> 103.36 GFLOPs | Improved |
Closest papers by meaning
- Beyond Single-Modal Boundary: Cross-Modal Anomaly Detection through Visual Prototype and Harmonization (CVPR 2025 · similarity 0.92)
- Rethinking Reverse Distillation for Multi-Modal Anomaly Detection (AAAI 2024 · similarity 0.89)
- Multimodal Industrial Anomaly Detection by Crossmodal Feature Mapping (CVPR 2024 · similarity 0.87)
- Exploring Intrinsic Normal Prototypes within a Single Image for Universal Anomaly Detection (CVPR 2025 · similarity 0.86)
- Align3D-AD: Cross-Modal Feature Alignment and Dual-Prompt Learning for Zero-shot 3D Anomaly Detection (NeurIPS 2026 · similarity 0.86)
- Real-IAD D3: A Real-World 2D/Pseudo-3D/3D Dataset for Industrial Anomaly Detection (CVPR 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.
