---
title: Open-HumanWAM-50K
canonical_url: "https://www.modelscope.cn/datasets/DAGroup-PKU/Open-HumanWAM-50K"
md_url: "https://www.modelscope.cn/datasets/DAGroup-PKU/Open-HumanWAM-50K.md"
repository: DAGroup-PKU/Open-HumanWAM-50K
last_updated: 2026-11-16
license: apache-2.0
storage_size: "1.9 TB"
downloads: 846568
stars: 0
---

# Open-HumanWAM-50K

> Open-HumanWAM-50K - DAGroup-PKU 在 ModelScope 开源的数据集。An Open Recipe for Scaling Human and Robot Experience for Embodied Foundation Models

DAGroup-PKU/Open-HumanWAM-50K 是 ModelScope 魔搭社区上的数据集，存储大小 1.9 TB，采用 apache-2.0 许可。

- **Repository**: DAGroup-PKU/Open-HumanWAM-50K
- **License**: apache-2.0
- **Storage size**: 1.9 TB
- **Downloads**: 846568
- **Stars**: 0
- **Last updated**: 2026-11-16

Source: https://www.modelscope.cn/datasets/DAGroup-PKU/Open-HumanWAM-50K

---

<!-- project-links:start -->
# Open-HumanWAM

**An Open Recipe for Scaling Human and Robot Experience for Embodied Foundation Models**

[Report](https://arxiv.org/abs/2611.06817) · [Code](https://github.com/DAGroup-PKU/HumanNet/tree/Open-HumanWAM) · [Project page](https://dagroup-pku.github.io/open-humanwam/) · [Models](https://huggingface.co/collections/DAGroup-PKU/open-humanwam) · [HF Paper](https://huggingface.co/papers/2611.06817) · [Citation](#citation)
<!-- project-links:end -->

![Open-HumanWAM teaser: the 50,000-hour corpus, human and robot scaling studies, and downstream robot results](assets/teaser.png)

*Open-HumanWAM overview: scaling human and robot experience for embodied foundation models.*

## Open-HumanWAM-50K

**50,000 hours of egocentric human manipulation video with hand-motion and
natural-language annotations for robot learning.**

Open-HumanWAM-50K is the human-data corpus accompanying
*Open-HumanWAM: An Open Recipe for Scaling Human and Robot Experience for
Embodied Foundation Models*. It brings together **HumanNet, EgoSuite-Open100K,
EgoDex, Gen-HumanEgo, and EgoVerse**, covering household, commercial-service,
and industrial manipulation.

The corpus is curated for hand visibility, motion stability, and manipulation
activity. Cross-source alignment standardizes camera coordinates, wrist axes,
and hand/gripper openness into a unified bimanual end-effector representation.
The 50,000-hour selection comprises **35.9 million selected clips** drawn from
**31,396,799 physical episodes**; multiple clips can reference the same episode.

![Open-HumanWAM-50K dataset construction: multi-source egocentric video, data curation, and cross-source action alignment](assets/data_overview.png)

*From diverse egocentric recordings to curated video, language, and hand-motion data for robot learning.*

## Format and organization

Data is stored in **LeRobot v3** format: compressed MP4 video, Parquet
action/state data, captions, episode metadata, and source provenance.
Captions are included; precomputed text embeddings are not.

| Directory | Source grouping |
| --- | --- |
| `egostandard/` | EgoStandard from EgoSuite-Open100K |
| `egopro/` | EgoPro from EgoSuite-Open100K |
| `egodex/` | EgoDex |
| `egoverse/` | EgoVerse, corresponding to the selected Mecka data |
| `humannet/` | HumanNet, Gen-HumanEgo, and other retained sources |

Each directory contains `part-*` folders. Each part is an independent LeRobot v3
dataset with `data/`, `videos/`, and `meta/`; `provenance/` preserves the original
source identity. See [collection.json](collection.json) for collection metadata.

## Choose a training subset

Seven subset configurations support experiments at different scales:
**100h, 500h, 1000h, 5000h, 10000h, 20500h, and 50000h**.
Each preserves the corresponding historical training selection, including the
supplemental episodes referenced by the smaller subsets.

**[Subset guide](subsets/README.md)** · [Subset manifest](subsets/manifest.json)

Each `subsets/<size>.parquet` file is a self-contained clip index with these fields:

| Field | Meaning |
| --- | --- |
| `dataset_path` | Relative path to a LeRobot part |
| `episode_index` | Episode identifier within that part |
| `start_frame`, `end_frame` | Episode-local frame interval, with an exclusive end |
| `source` | Original data source |
| `task` | Natural-language annotation for the clip |
| `hand` | Hand annotation used by the training selection |

The Hugging Face configurations load these **clip indices**. Download the
referenced LeRobot parts separately to access the video and action/state data.

## Access and training

To access the data, sign in to Hugging Face and complete the access request form.
Requests are approved automatically. For scripted downloads, authenticate with
a token from the approved account, for example with `hf auth login`.

For model setup and training instructions, see the
[Open-HumanWAM code repository](https://github.com/DAGroup-PKU/HumanNet/tree/Open-HumanWAM).

## License

This project's original contributions are released under the
[Apache License 2.0](LICENSE). The underlying third-party datasets remain subject
to their original licenses and terms. This release does not replace or expand
the permissions granted by those source datasets; applicable source terms
continue to apply to their content.

## Citation

If you use Open-HumanWAM-50K in your research, please cite our
[technical report](https://arxiv.org/abs/2611.06817):

```bibtex
@misc{ma2026openhumanwam,
  title = {{Open-HumanWAM}: An Open Recipe for Scaling Human and Robot Experience for Embodied Foundation Models},
  author = {Ma, Juncheng and Li, Jiaxin and Tu, Jiankai and others},
  year = {2026},
  eprint = {2611.06817},
  archivePrefix = {arXiv},
  primaryClass = {cs.CV},
  url = {https://arxiv.org/abs/2611.06817}
}
```
