|
|
--- |
|
|
task_categories: |
|
|
- image-to-video |
|
|
license: cc-by-4.0 |
|
|
language: |
|
|
- en |
|
|
tags: |
|
|
- panoramic |
|
|
- video-generation |
|
|
- motion-control |
|
|
- 360-degree |
|
|
- optical-flow |
|
|
- computer-vision |
|
|
- diffusion |
|
|
--- |
|
|
|
|
|
# PanFlow Dataset |
|
|
|
|
|
The PanFlow dataset supports the research presented in the paper **[PanFlow: Decoupled Motion Control for Panoramic Video Generation](https://huggingface.co/papers/2512.00832)**. |
|
|
|
|
|
PanFlow is a novel framework for controllable 360° panoramic video generation that decouples motion input into two interpretable components: rotation flow and derotated flow. This dataset is a large-scale, motion-rich panoramic video dataset with frame-level pose and optical flow annotations, curated to enable precise motion control, produce loop-consistent panoramas, and support applications such as motion transfer and panoramic video editing. |
|
|
|
|
|
**Paper:** [https://huggingface.co/papers/2512.00832](https://huggingface.co/papers/2512.00832) |
|
|
**Code:** [https://github.com/chengzhag/PanFlow](https://github.com/chengzhag/PanFlow) |
|
|
**Video Overview:** [https://www.youtube.com/watch?v=sFTWwlHjNtg](https://www.youtube.com/watch?v=sFTWwlHjNtg) |
|
|
|
|
|
<p align="center"> |
|
|
<img src="images/flow.png" alt="flow" width="400"> |
|
|
</p> |
|
|
|
|
|
By conditioning diffusion on spherical-warped motion noise, PanFlow enables precise motion control, produces loop-consistent panoramas, and supports applications such as motion transfer: |
|
|
|
|
|
<p align="center"> |
|
|
<img src="images/transfer.gif" alt="flow" width="860"> |
|
|
</p> |
|
|
|
|
|
and panoramic video editing: |
|
|
|
|
|
<p align="center"> |
|
|
<img src="images/editing.gif" alt="flow" width="860"> |
|
|
</p> |
|
|
|
|
|
## Dataset Structure and Details |
|
|
|
|
|
The PanFlow dataset provides camera pose annotations for 300k clips. It also includes pre-generated latent and noise cache for a filtered subset to speed up training. |
|
|
|
|
|
The underlying video data is derived from the [360-1M dataset](https://github.com/MattWallingford/360-1M), which consists of YouTube videos licensed under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/). We provide a 720P version on [360-1M-720P](https://huggingface.co/datasets/chengzhag/360-1M-720P). |
|
|
|
|
|
## Citation |
|
|
|
|
|
If you use the PanFlow dataset in your research, please cite the original paper: |
|
|
|
|
|
```bibtex |
|
|
@inproceedings{zhang2025panflow, |
|
|
title={PanFlow: Decoupled Motion Control for Panoramic Video Generation}, |
|
|
author={Zhang, Cheng and Liang, Hanwen and Chen, Donny Y and Wu, Qianyi and Plataniotis, Konstantinos N and Gambardella, Camilo Cruz and Cai, Jianfei}, |
|
|
booktitle={Proceedings of the AAAI Conference on Artificial Intelligence}, |
|
|
year={2026} |
|
|
} |
|
|
``` |
