JoPano: Unified Panorama Generation via Joint Modeling
Wancheng Feng^1,3*, Chen An^1,2*, Zhenliang He^1✉, Meina Kan^1,2, Shiguang Shan^1,2, Lukun Wang^3
1State Key Lab of AI Safety, Institute of Computing Technology, CAS, China
²University of Chinese Academy of Sciences (CAS), China
³Shandong University of Science and Technology, China

We propose a joint-face panorama (JoPano) generation approach that unifies the two core tasks within a DiT-based model. To transfer the rich generative capabilities of existing DiT backbones learned from natural images to the panorama domain, we propose a Joint-Face Adapter built on the cubemap representation of panoramas, which enables a pretrained DiT to jointly model and generate different views of a panorama. We further apply Poisson Blending to reduce seam inconsistencies that often appear at the boundaries between cube faces.

• [2025-12-07]: The arXiv paper, project page are released.

Comming soon.

This project is built upon Sana. Thanks for their great work!

If you find this project helpful, please consider citing:

@article{feng2025jopano,
  title={JoPano: Unified Panorama Generation via Joint Modeling},
  author={Feng, Wancheng and An, Chen and He, Zhenliang and Kan, Meina and Shan, Shiguang and Wang, Lukun},
  journal={arXiv:2512.06885},
  year={2025}
}