We introduce Shape-for-Motion, a 3D-aware video editing framework to support precise and consistent video object manipulation by reconstructing an editable 3D mesh to serve as control signals for video generation.

image/jpeg

πŸ”— BibTeX

If you find Shape-for-Motion useful for your research and applications, please cite using this BibTeX:

@article{liu2025shape,
  title={Shape-for-Motion: Precise and Consistent Video Editing with 3D Proxy},
  author={Liu, Yuhao and Wang, Tengfei and Liu, Fang and Wang, Zhenwei and Lau, Rynson WH},
  journal={arXiv preprint arXiv:2506.22432},
  year={2025}
}

Acknowledgements

We would like to thank DG-Mesh, Deformable 3DGS, and Diffusers for their open research and exploration.

Downloads last month

-

Downloads are not tracked for this model. How to track
Safetensors
Model size
682M params
Tensor type
F32
Β·
Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support

Model tree for LeoLau/Shape-for-Motion

Finetuned
(1)
this model