BASKET-Multiview
收藏资源简介:
我们引入了BASKET-Multiview数据集,这是一个表示常见篮球比赛的合成场景集合。我们提供了全面的注释,包括校准的相机、动画、RGB图像、分割掩码、深度图、表面法线、SMPL网格和动画。所有场景都以1080p/4K和30 fps渲染。
We introduce the BASKET-Multiview dataset, a collection of synthetic scenes representing typical basketball game scenarios. We provide comprehensive annotations including calibrated cameras, animations, RGB images, segmentation masks, depth maps, surface normals, SMPL meshes, and animations. All scenes are rendered at 1080p/4K resolution and 30 frames per second.
BASKET-Multiview 数据集概述
数据集简介
BASKET-Multiview 是一个合成篮球比赛场景数据集,包含常见的篮球战术场景。该数据集提供全面的标注信息,用于支持动态3D场景重建研究。
数据内容
- 数据类型:多视角视频流
- 场景内容:篮球比赛常见战术场景
- 渲染规格:1080p/4K分辨率,30帧/秒
- 标注信息:
- 标定相机参数
- 动画数据
- RGB图像
- 分割掩码
- 深度图
- 表面法线
- SMPL网格和动画
数据获取
数据集下载说明位于:https://humansensinglab.github.io/basket-multiview/data.html#download_section 提供可视化脚本帮助理解数据集约定
注意事项
SMPL网格是通过拟合真实骨骼获得的,可能不完美。由于许可限制,无法重新分发原始角色资产,研究人员可从原始来源获取完美网格:https://www.fab.com/listings/2eebf211-df72-4daf-a8c4-24bfadba9e7a
相关挑战
发布BASKET-Multiview重建挑战,评估多视角观察动态场景重建方法。详细提交说明和评估政策参见竞赛页面:https://challenge.shannon.humansensing.cs.cmu.edu/web/challenges/challenge-page/21/overview
引用信息
如需使用本项目,请引用:
@article{10.1145/3731214, author = {Huang, Junkai and Subhajyoti Mallick, Saswat and Amat, Alejandro and Ruiz Olle, Marc and Mosella-Montoro, Albert and Kerbl, Bernhard and Vicente Carrasco, Francisco and De la Torre, Fernando}, title = {Echoes of the Coliseum: Towards 3D Live streaming of Sports Events}, year = {2025}, issue_date = {August 2025}, publisher = {Association for Computing Machinery}, address = {New York, NY, USA}, volume = {44}, number = {4}, issn = {0730-0301}, url = {https://doi.org/10.1145/3731214}, doi = {10.1145/3731214}, journal = {ACM Trans. Graph.}, month = jul, articleno = {46}, numpages = {17}, keywords = {sports events, human-centric events, gaussian splatting, real-time 3D reconstruction} }




