nielsr HF Staff commited on
Commit
1ce9e20
·
verified ·
1 Parent(s): 6db2fa3

Improve model card and metadata

Browse files

Hi! I'm Niels from the Hugging Face community team.

This pull request improves the model card for RA-Monitor by:
- Adding the `image-text-to-text` pipeline tag for better discoverability.
- Specifying `library_name: transformers` based on the model configuration.
- Including a description of the M3-AD framework and RA-Monitor mechanism.
- Providing links to the research paper and official GitHub repository.

These changes help users understand the model's purpose and how to cite the work.

Files changed (1) hide show
  1. README.md +32 -3
README.md CHANGED
@@ -1,3 +1,32 @@
1
- ---
2
- license: apache-2.0
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ pipeline_tag: image-text-to-text
4
+ library_name: transformers
5
+ ---
6
+
7
+ # M3-AD: RA-Monitor
8
+
9
+ This repository contains the model weights for **RA-Monitor**, a unified reflection-aware multimodal framework for industrial anomaly detection. RA-Monitor is part of the M3-AD framework presented in the paper [M3-AD: Reflection-aware Multi-modal, Multi-category, and Multi-dimensional Benchmark and Framework for Industrial Anomaly Detection](https://huggingface.co/papers/2603.00055).
10
+
11
+ ## Model Description
12
+ RA-Monitor addresses the issue where multimodal large language models (MLLMs) produce high-confidence but unreliable decisions in complex industrial scenarios. It introduces a reflection-aware mechanism that models reflection as a learnable decision revision process. This allows the model to perform controlled self-correction when initial judgments are unreliable, significantly improving anomaly type recognition and spatial localization.
13
+
14
+ The framework is built upon:
15
+ - **RA-Monitor**: A mechanism that equips pre-trained models with thinking and reflective abilities.
16
+ - **M3-AD-FT**: A dataset designed for reflection-aligned fine-tuning.
17
+ - **M3-AD-Bench**: A benchmark for systematic cross-category evaluation of industrial anomaly detection.
18
+
19
+ ## Resources
20
+ - **Paper:** [M3-AD: Reflection-aware Multi-modal, Multi-category, and Multi-dimensional Benchmark and Framework for Industrial Anomaly Detection](https://huggingface.co/papers/2603.00055)
21
+ - **GitHub Repository:** [Yanhui-Lee/M3-AD](https://github.com/Yanhui-Lee/M3-AD)
22
+
23
+ ## Citation
24
+ If you find this work useful, please cite the following paper:
25
+ ```bibtex
26
+ @article{m3ad2026,
27
+ title={M3-AD: Reflection-aware Multi-modal, Multi-category, and Multi-dimensional Benchmark and Framework for Industrial Anomaly Detection},
28
+ author={Li, Yanhui and others},
29
+ journal={arXiv preprint arXiv:2603.00055},
30
+ year={2026}
31
+ }
32
+ ```