BFSM 数据集
收藏资源简介:
BFSM 数据集是一个包含超过 200 个样本的数据集,包括正常病例和罕见的颅面条件。每个病例都包含一个基于 CT 的头骨、一个基于 CT 的面部和一个高保真的纹理面部扫描。该数据集是迄今为止同类数据集中最为全面的,旨在用于远程诊断、手术计划、医学教育和基于物理的面部模拟等应用。数据集创建过程中,我们与经验丰富的临床医生紧密合作,确保数据的解剖准确性。BFSM 数据集可用于远程医疗和手术计划等应用,展示了其在促进医患沟通和支持更明智的临床决策方面的潜力。
The BFSM dataset is a collection of over 200 samples covering both normal cases and rare craniofacial conditions. Each case includes a CT-based skull model, a CT-based facial model, and a high-fidelity textured facial scan. This is the most comprehensive dataset of its kind to date, designed for applications such as remote diagnosis, surgical planning, medical education, and physics-based facial simulation. During its development, we collaborated closely with experienced clinicians to ensure the anatomical accuracy of the dataset. The BFSM dataset can be applied in scenarios such as telemedicine and surgical planning, demonstrating its potential to facilitate doctor-patient communication and support more informed clinical decision-making.
BFSM: 3D双向人脸-颅骨可变形模型
数据集概述
BFSM是一个3D双向人脸-颅骨可变形模型,实现了人脸与颅骨之间的双向变形映射。
作者信息
- Zidu Wang
- Meng Xu
- Miao Xu
- Hengyuan Ma
- Jiankuo Zhao
- Xutao Li
- Xiangyu Zhu
- Zhen Lei
技术实现
- 官方实现代码库:https://github.com/wang-zidu/BFSM
- 相关论文:https://arxiv.org/abs/2509.24577
使用限制
⚠️ 免责声明:本项目仅用于研究和测试目的。禁止用于医疗或临床应用,严禁商业用途。
引用信息
如需在研究中使用本工作,请引用以下论文:
@misc{wang2025bfsm, title={BFSM: 3D Bidirectional Face-Skull Morphable Model}, author={Zidu Wang and Meng Xu and Miao Xu and Hengyuan Ma and Jiankuo Zhao and Xutao Li and Xiangyu Zhu and Zhen Lei}, year={2025}, eprint={2509.24577}, archivePrefix={arXiv}, primaryClass={cs.CV}, url={https://arxiv.org/abs/2509.24577}, }




