Research map / Label-efficient and robust learning / OOD and anomaly detection
ReMP-AD: Retrieval-enhanced Multi-modal Prompt Fusion for Few-Shot Industrial Visual Anomaly Detection
Hongchi Ma, Guanglei Yang, Debin Zhao, Yanli Ji, Wangmeng Zuo
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 Global Prototype Retrieval (ablation) | ReMP-AD baseline · VisA (4-shot) | P-F1: 39.7 -> 45.4 | Improved |
| add Global Prototype Retrieval (ablation) | ReMP-AD baseline · VisA (4-shot) | I-AUC: 91.5 -> 94.0 | Improved |
| add Global Prototype Retrieval (ablation) | ReMP-AD baseline · VisA (4-shot) | I-F1: 87.4 -> 89.3 | Improved |
| add Class Token Balancing on top of GPR (ablation) | ReMP-AD · VisA (4-shot) | P-F1: 45.4 -> 46.9 | Improved |
| add Class Token Balancing on top of GPR (ablation) | ReMP-AD · VisA (4-shot) | I-AUC: 94.0 -> 94.0 | No clear effect |
| add Vision-Language Prior Fusion on top of ICTR (ablation) | ReMP-AD · VisA (4-shot) | P-F1: 46.9 -> 48.0 | Improved |
| add CTB (4-shot reference weighting) (ablation) | ReMP-AD · VisA (4-shot) | P-F1: 46.6 -> 48.0 | Improved |
| add CTB (1-shot) (ablation) | ReMP-AD · VisA (1-shot) | I-AUC: 91.2 -> 90.9 | Hurt |
| add CTB (16-shot) (ablation) | ReMP-AD · VisA (16-shot) | P-F1: 49.4 -> 50.3 | Improved |
| add ICTR module to PromptAD | PromptAD · VisA (4-shot) | P-F1: 36.5 -> 37.4 | Improved |
| add ICTR module to PromptAD | PromptAD · MVTec-AD (1-shot) | P-AUC: 95.4 -> 94.9 | Hurt |
| add VLPF image encoder (mask attention) to MuSc | MuSc · MVTec-AD (many-shot) | P-F1: 62.2 -> 63.0 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- PromptAD — PromptAD: Learning Prompts with only Normal Samples for Few-Shot Anomaly Detection (CVPR 2024) · better on 6 of 6 shared results
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
- AnomalyGPT: Detecting Industrial Anomalies Using Large Vision-Language Models (AAAI 2024 · similarity 0.90)
- FIND: Few-Shot Anomaly Inspection with Normal-Only Multi-Modal Data (ICCV 2025 · similarity 0.90)
- PromptAD: Learning Prompts with only Normal Samples for Few-Shot Anomaly Detection (CVPR 2024 · similarity 0.90)
- Bidirectional Multimodal Prompt Learning with Scale-Aware Training for Few-Shot Multi-Class Anomaly Detection (CVPR 2026 · similarity 0.90)
- Multimodal Industrial Anomaly Detection by Crossmodal Feature Mapping (CVPR 2024 · similarity 0.89)
- MuSc: Zero-Shot Industrial Anomaly Classification and Segmentation with Mutual Scoring of the Unlabeled Images (ICLR 2024 · similarity 0.88)
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.
