[ECCV 2026] PhyGDPO: Physics-Aware Groupwise Direct Preference Optimization for Physically Consistent Text-to-Video Generation

Model Description

Here are two models based on Wan2.1-1.3B and Wan2.1-14B for physically consistent text-to-video generation. The two models are post-trained by our re-implemented PhyGDPO algorithm on public datasets.

(1) A gymnast drops from the parallel bars and lands safely on the mat below.
Ours	OpenAI Sora 2	Google Veo 3.1
(2) A soccer player runs, plants their foot, and drop-kicks a soccer ball high into the air, with the ball visibly arcing through the air.
Ours	OpenAI Sora 2	Google Veo 3.1
(3) A person wearing a helmet performs a handspring over a platform.
Ours	OpenAI Sora 2	Google Veo 3.1
(4) A person plays squash on an indoor court.
Ours	OpenAI Sora 2	Google Veo 3.1

Github Code Link

This dataset is intended to be used together with our code. Please refer to the GitHub repository below for more detailed instructions.

https://github.com/caiyuanhao1998/Open-PhyGDPO

Huggingface Dataset Link

We also release three models based on Wan2.1-1.3B and Wan2.1-14B in the following link:

https://huggingface.co/datasets/CaiYuanhao/PhyGDPO

Project Page Link

For more video customization results, please refer to our project page:

https://caiyuanhao1998.github.io/project/PhyGDPO/

Arxiv Paper Link

For more technical details, please refer to our ECCV 2026 paper:

https://arxiv.org/abs/2512.24551

知乎中文解读

https://zhuanlan.zhihu.com/p/2060200098131350370

Citation

If you find our code, data, and models useful, please consider citing our paper:

  @inproceedings{phygdpo,
  title={PhyGDPO: Physics-Aware Groupwise Direct Preference Optimization for Physically Consistent Text-to-Video Generation},
  author={Cai, Yuanhao and Li, Kunpeng and Jia, Menglin and Wang, Jialiang and Sun, Junzhe and Liang, Feng and Chen, Weifeng and Juefei-Xu, Felix and Wang, Chu and Thabet, Ali and Dai, Xiaoliang and Ju, Xuan and Yuille, Alan and Hou, Ji},
  booktitle={ECCV},
  year={2026}
}