README.md

---
license: cc-by-nc-sa-4.0
base_model:
- Ultralytics/YOLO11
tags:
- chemistry
---

# Molecule Detection YOLO in MolParser

From paper: "*MolParser: End-to-end Visual Recognition of Molecule Structures in the Wild*" (ICCV2025 Accept)

[Arxiv Paper](https://arxiv.org/abs/2411.11098) | [Huggingface Dataset](https://huggingface.co/datasets/AI4Industry/MolParser-7M) | [OCSR Demo](https://ocsr.dp.tech/) | [MolDetv2](https://huggingface.co/UniParser/MolDetv2)

We provide several [ultralytics YOLO11](https://github.com/ultralytics/ultralytics) weights for molecule detection with different size & input resolution.

![image/png](https://cdn-uploads.huggingface.co/production/uploads/65f7f16fb6941db5c2e7c4bf/7oWPoPxuEXSangDWnJ7mv.png)


## 1⃣️ [MolDet-General] General molecule structure detection models

`moldet_yolo11[size]_640_general.pt`

YOLO11 weights trained on 35k human annotated image crops and 100k generated images

* 640x640 input resolution
* support handwritten molecules
* **multiscale input** (inputs can be single/multiple molecular cutouts, reaction or table cutouts, or single-page PDF images)

<span style='color:gray'>Warning: For single-molecule input (used as a classification model), appropriate padding can be added to enhance the performance.</span>

Result in private testing:
| Model Size | mAP50 | mAP50-95 | Speed (T4 TensorRT10) |
| ---- | ----- | -------- | ----- |
| n | 0.9581 | 0.8524 | 1.5 ± 0.0 ms |
| s | 0.9652 | 0.8704 | 2.5 ± 0.1 ms |
| m | 0.9686 | 0.8736 | 4.7 ± 0.1 ms |
| l | **0.9891** | **0.9028** | 6.2 ± 0.1 ms |

usage:
```python
from ultralytics import YOLO
model = YOLO("moldet_yolo11l_640_general.pt")
model.predict("path/to/image.png", save=True, imgsz=640, conf=0.5)
```

## 2⃣️ [MolDet-Doc] PDF molecule structure detection models

`moldet_yolo11[size]_960_doc.pt`

YOLO11 weights trained on 26k human annotated PDF pages (patents, papers, and books)

* 960x960 input resolution
* prefer **single page PDF image** input
* better in small molecule detection

<span style='color:gray'>Warning: It is recommended to use MuPDF to render PDF pages at more than 144dpi.</span>


Result in private testing:
| Model Size | mAP50 | mAP50-95 | Speed (T4 TensorRT10) |
| ---- | ----- | -------- | ----- |
| n | 0.9871 | 0.8732 | 3.1 ± 0.0 ms |
| s | 0.9851 | 0.8824 | 5.5 ± 0.1 ms |
| m | 0.9867 | 0.8917 | 9.9 ± 0.2 ms |
| l | **0.9913** | **0.9011** | 13.1 ± 0.3 ms |

usage:
```python
from ultralytics import YOLO
model = YOLO("moldet_yolo11l_960_doc.pt")
model.predict("path/to/pdf_page_image.png", save=True, imgsz=960, conf=0.5)
```

## 📜 License

MolDet & MolDetv2 model weights are provided for **non-commercial use only**. 

For commercial use, please contact: [fangxi@dp.tech](mailto:fangxi@dp.tech) or add a discussion in HuggingFace.

## 📖 Citation

If you use this model in your work, please cite:

```
@inproceedings{fang2025molparser,
  title={Molparser: End-to-end visual recognition of molecule structures in the wild},
  author={Fang, Xi and Wang, Jiankun and Cai, Xiaochen and Chen, Shangqian and Yang, Shuwen and Tao, Haoyi and Wang, Nan and Yao, Lin and Zhang, Linfeng and Ke, Guolin},
  booktitle={Proceedings of the IEEE/CVF International Conference on Computer Vision},
  pages={24528--24538},
  year={2025}
}
```