A dataset for tooth structured instance segmentation of dental panoramic X-ray.
The dataset will be available for download . Stay tuned for updates.
The dataset is organized as follows:
TSD/
├── images
│ ├── img_001.png
│ ├── img_002.png
│ ├── img_003.png
│ └── ...
├── train/val/test_json_files/
│ ├── info
│ ├── license
│ ├── images_reference
│ ├── annotations/
│ │ ├── instance 1
│ │ ├── instance 2
│ │ │ ├── segmentation (list)
│ │ │ │ ├── instance_polygons
│ │ │ │ ├── dentin_polygons
│ │ │ │ ├── pulp_polygons
│ │ │ │ ├── materials_polygons
│ │ │ │ └── decay_polygons
│ │ │ ├── image_id (image ref of instance)
│ │ │ ├── instance_bounding_box
│ │ │ └── instance_id
│ │ └── ... instance n
Each segmentation annotation consists of a list of 5 items, with all polygons represented by sequences of points in the format (x1, y1, x2, y2, ...).
To organize the structured instance of a tooth, arrange the polygons in the specified order.
Our research paper provides detailed information about the dataset, the methodology used for annotation, and the results of our experiments.
If you use this dataset in your research, please cite our paper as follows:
@article{wang2024dsis,
title={DSIS-DPR: Structured Instance Segmentation and Diffusion Prior Refinement for Dental Anatomy Learning},
author={Wang, Xianyun and Wang, Linhong and Yang, Zhenchen and Zhou, Jiacong and Zheng, Yuchen and Chen, Feng and Hong, Richang and Yu, Jun and Yang, Fan},
journal={IEEE Transactions on Multimedia},
year={2024},
publisher={IEEE}
}
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent