USC-PSI-Lab/Humanoid-X

---
task_categories:
- robotics
---

<div align="center">
<h1> <img src="assets/icon.png" width="50" /> Humanoid-X </h1>
</div>
<h5 align="center">
<a href="https://usc-gvl.github.io/UH-1/">🌐 Homepage</a> | <a href="https://huggingface.co/datasets/USC-GVL/Humanoid-X">⛁ Dataset</a> | <a href="https://huggingface.co/USC-GVL/UH-1">🤗 Models</a> | <a href="https://arxiv.org/abs/2412.14172">📑 Paper</a> | <a href="https://github.com/sihengz02/UH-1">💻 Code</a>
</h5>

This repo contains the officail dataset for the paper "[Learning from Massive Human Videos for Universal Humanoid Pose Control](https://arxiv.org/abs/2412.14172)"
If you like our project, please give us a star ⭐ on GitHub for latest update.


- In this repo, we fully release the text desciption data `texts.zip`, humanoid keypoints data `humanoid_keypoint.zip`, and humanoid actions data `humanoid_action.zip`.

- We only release part of the human poses data (charades subset, kinetics700 subset, and youtube subset) `human_pose.zip` due to license issues. Instead, we provide instructions on how to obtain other parts of human poses data: [HumanML3D/AMASS](https://github.com/EricGuo5513/HumanML3D), [Motion-X](https://github.com/IDEA-Research/Motion-X?tab=readme-ov-file#-dataset-download).

- We release the train, test, and valid set split as `train.txt`, `test.txt`, and `val.txt`.

- We will not release the original Internet videos to protect copyright.

# Dataset Statistics


# Dataset Collection Pipeline


# Citation

If you find our work helpful, please cite us:

```bibtex
@article{mao2024learning,
title={Learning from Massive Human Videos for Universal Humanoid Pose Control},
author={Mao, Jiageng and Zhao, Siheng and Song, Siqi and Shi, Tianheng and Ye, Junjie and Zhang, Mingtong and Geng, Haoran and Malik, Jitendra and Guizilini, Vitor and Wang, Yue},
journal={arXiv preprint arXiv:2412.14172},
year={2024}
}
```

task_categories:
- 任务类别：机器人学（robotics）

<div align="center">
<h1> <img src="assets/icon.png" width="50" /> 人形-X（Humanoid-X）</h1>
</div>
<h5 align="center">
<a href="https://usc-gvl.github.io/UH-1/">🌐 主页</a> | <a href="https://huggingface.co/datasets/USC-GVL/Humanoid-X">⛁ 数据集</a> | <a href="https://huggingface.co/USC-GVL/UH-1">🤗 模型</a> | <a href="https://arxiv.org/abs/2412.14172">📑 论文</a> | <a href="https://github.com/sihengz02/UH-1">💻 代码</a>
</h5>

本仓库承载了论文《面向通用人形姿态控制的海量人类视频学习》（Learning from Massive Human Videos for Universal Humanoid Pose Control，https://arxiv.org/abs/2412.14172）的官方数据集。若您对本项目感兴趣，欢迎前往GitHub为我们点亮Star ⭐ 以获取最新更新。


- 本仓库完整开放文本描述数据集`texts.zip`、人形关键点（humanoid keypoints）数据集`humanoid_keypoint.zip`以及人形动作数据集`humanoid_action.zip`。
- 受版权许可限制，我们仅开放部分人体姿态数据集（Charades子集、Kinetics700子集以及YouTube子集）`human_pose.zip`。针对剩余部分的人体姿态数据集，我们提供了获取指南：[HumanML3D/AMASS](https://github.com/EricGuo5513/HumanML3D)、[Motion-X](https://github.com/IDEA-Research/Motion-X?tab=readme-ov-file#-dataset-download)。
- 我们将训练集、测试集与验证集的划分信息分别存储于`train.txt`、`test.txt`与`val.txt`文件中。
- 为保护版权，我们不会公开原始网络视频素材。

# 数据集统计信息


# 数据集采集流程


# 引用声明

若您的工作受益于本项目，请引用我们的成果：

bibtex
@article{mao2024learning,
title={Learning from Massive Human Videos for Universal Humanoid Pose Control},
author={Mao, Jiageng and Zhao, Siheng and Song, Siqi and Shi, Tianheng and Ye, Junjie and Zhang, Mingtong and Geng, Haoran and Malik, Jitendra and Guizilini, Vitor and Wang, Yue},
journal={arXiv preprint arXiv:2412.14172},
year={2024}
}