lakshhhchawla/GTA-Human

---
license: other
license_name: s-lab-license
license_link: LICENSE
language:
- en
tags:
- Human Pose and Shape Estimation
- 3D Human Recovery
- Parametric Humans
- Synthetic Data
pretty_name: GTA-Human
size_categories:
- 100B<n<1T
---

# Playing for 3D Human Recovery (TPAMI 2024)
- [Homepage](https://caizhongang.com/projects/GTA-Human/)
- [Toolbox](https://github.com/caizhongang/gta-human_toolbox/tree/main)
- [Paper](https://arxiv.org/abs/2110.07588)

## Updates

- [2024-10-02] GTA-Human datasets are now available on HuggingFace!
- [2024-09-19] Release of GTA-Human II Dataset
- [2022-07-08] Release of GTA-Human Dataset on MMHuman3D

## Datasets

Please click on the dataset name for download links and visualization instructions.

| Features | [GTA-Human](https://caizhongang.com/projects/GTA-Human/gta-human.html) | [GTA-Human II](https://caizhongang.com/projects/GTA-Human/gta-human_v2.html) |
| :------------------------ | :-------: | :-------: |
| Num of Scenes | 20,005 | 10,224 |
| Num of Person Sequences | 20,005 | 35,352 |
| Color Images | Yes | Yes |
| 3D BBox & Point Cloud | No | Yes |
| Parametric Model | SMPL | SMPL-X |
| Num of Persons per Scene | 1 | 1-6 |

## Citation
```text
@ARTICLE{10652891,
author={Cai, Zhongang and Zhang, Mingyuan and Ren, Jiawei and Wei, Chen and Ren, Daxuan and Lin, Zhengyu and Zhao, Haiyu and Yang, Lei and Loy, Chen Change and Liu, Ziwei},
journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},
title={Playing for 3D Human Recovery},
year={2024},
volume={},
number={},
pages={1-12},
keywords={Three-dimensional displays;Annotations;Synthetic data;Shape;Training;Parametric statistics;Solid modeling;Human Pose and Shape Estimation;3D Human Recovery;Parametric Humans;Synthetic Data;Dataset},
doi={10.1109/TPAMI.2024.3450537}
}
```

许可证：其他
许可证名称：s-lab-license
许可证链接：LICENSE
语言：
- 英语
标签：
- 人体姿态与形状估计（Human Pose and Shape Estimation）
- 三维人体恢复（3D Human Recovery）
- 参数化人体模型（Parametric Humans）
- 合成数据（Synthetic Data）
数据集名称：GTA-Human
数据规模分类：
- 1000亿 < 数据量 < 1万亿

# 面向三维人体恢复的研究（TPAMI 2024）
- [项目主页](https://caizhongang.com/projects/GTA-Human/)
- [工具箱](https://github.com/caizhongang/gta-human_toolbox/tree/main)
- [论文](https://arxiv.org/abs/2110.07588)

## 更新日志
- [2024-10-02] GTA-Human 数据集现已上架 HuggingFace！
- [2024-09-19] GTA-Human II 数据集正式发布
- [2022-07-08] GTA-Human 数据集于 MMHuman3D 平台发布

## 数据集
请点击数据集名称获取下载链接与可视化说明。

| 特征项 | GTA-Human | GTA-Human II |
| :------------------------ | :-------: | :-------: |
| 场景总数 | 20005 | 10224 |
| 人体序列总数 | 20005 | 35352 |
| 彩色图像 | 支持 | 支持 |
| 三维边界框与点云数据 | 不支持 | 支持 |
| 参数化人体模型 | SMPL模型（SMPL） | SMPL-X模型（SMPL-X） |
| 单场景人数 | 1 | 1-6 |

## 引用格式
text
@ARTICLE{10652891,
author={Cai, Zhongang and Zhang, Mingyuan and Ren, Jiawei and Wei, Chen and Ren, Daxuan and Lin, Zhengyu and Zhao, Haiyu and Yang, Lei and Loy, Chen Change and Liu, Ziwei},
journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},
title={Playing for 3D Human Recovery},
year={2024},
volume={},
number={},
pages={1-12},
keywords={Three-dimensional displays;Annotations;Synthetic data;Shape;Training;Parametric statistics;Solid modeling;Human Pose and Shape Estimation;3D Human Recovery;Parametric Humans;Synthetic Data;Dataset},
doi={10.1109/TPAMI.2024.3450537}
}