Spherical-Nested Diffusion Model for Panoramic Image Outpainting

Xiancheng Sun, Senmao Ma, Shengxi Li, Mai Xu, Jingyuan Xia, Lai Jiang, Xin Deng, Jiali Wang
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:57614-57630, 2025.

Abstract

Panoramic image outpainting acts as a pivotal role in immersive content generation, allowing for seamless restoration and completion of panoramic content. Given the fact that the majority of generative outpainting solutions operates on planar images, existing methods for panoramic images address the sphere nature by soft regularisation during the end-to-end learning, which still fails to fully exploit the spherical content. In this paper, we set out the first attempt to impose the sphere nature in the design of diffusion model, such that the panoramic format is intrinsically ensured during the learning procedure, named as spherical-nested diffusion (SpND) model. This is achieved by employing spherical noise in the diffusion process to address the structural prior, together with a newly proposed spherical deformable convolution (SDC) module to intrinsically learn the panoramic knowledge. Upon this, the proposed method is effectively integrated into a pre-trained diffusion model, outperforming existing state-of-the-art methods for panoramic image outpainting. In particular, our SpND method reduces the FID values by more than 50% against the state-of-the-art PanoDiffusion method. Codes are publicly available at https://github.com/chronos123/SpND.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-sun25m, title = {Spherical-Nested Diffusion Model for Panoramic Image Outpainting}, author = {Sun, Xiancheng and Ma, Senmao and Li, Shengxi and Xu, Mai and Xia, Jingyuan and Jiang, Lai and Deng, Xin and Wang, Jiali}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {57614--57630}, year = {2025}, editor = {Singh, Aarti and Fazel, Maryam and Hsu, Daniel and Lacoste-Julien, Simon and Berkenkamp, Felix and Maharaj, Tegan and Wagstaff, Kiri and Zhu, Jerry}, volume = {267}, series = {Proceedings of Machine Learning Research}, month = {13--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v267/main/assets/sun25m/sun25m.pdf}, url = {https://proceedings.mlr.press/v267/sun25m.html}, abstract = {Panoramic image outpainting acts as a pivotal role in immersive content generation, allowing for seamless restoration and completion of panoramic content. Given the fact that the majority of generative outpainting solutions operates on planar images, existing methods for panoramic images address the sphere nature by soft regularisation during the end-to-end learning, which still fails to fully exploit the spherical content. In this paper, we set out the first attempt to impose the sphere nature in the design of diffusion model, such that the panoramic format is intrinsically ensured during the learning procedure, named as spherical-nested diffusion (SpND) model. This is achieved by employing spherical noise in the diffusion process to address the structural prior, together with a newly proposed spherical deformable convolution (SDC) module to intrinsically learn the panoramic knowledge. Upon this, the proposed method is effectively integrated into a pre-trained diffusion model, outperforming existing state-of-the-art methods for panoramic image outpainting. In particular, our SpND method reduces the FID values by more than 50% against the state-of-the-art PanoDiffusion method. Codes are publicly available at https://github.com/chronos123/SpND.} }
Endnote
%0 Conference Paper %T Spherical-Nested Diffusion Model for Panoramic Image Outpainting %A Xiancheng Sun %A Senmao Ma %A Shengxi Li %A Mai Xu %A Jingyuan Xia %A Lai Jiang %A Xin Deng %A Jiali Wang %B Proceedings of the 42nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Aarti Singh %E Maryam Fazel %E Daniel Hsu %E Simon Lacoste-Julien %E Felix Berkenkamp %E Tegan Maharaj %E Kiri Wagstaff %E Jerry Zhu %F pmlr-v267-sun25m %I PMLR %P 57614--57630 %U https://proceedings.mlr.press/v267/sun25m.html %V 267 %X Panoramic image outpainting acts as a pivotal role in immersive content generation, allowing for seamless restoration and completion of panoramic content. Given the fact that the majority of generative outpainting solutions operates on planar images, existing methods for panoramic images address the sphere nature by soft regularisation during the end-to-end learning, which still fails to fully exploit the spherical content. In this paper, we set out the first attempt to impose the sphere nature in the design of diffusion model, such that the panoramic format is intrinsically ensured during the learning procedure, named as spherical-nested diffusion (SpND) model. This is achieved by employing spherical noise in the diffusion process to address the structural prior, together with a newly proposed spherical deformable convolution (SDC) module to intrinsically learn the panoramic knowledge. Upon this, the proposed method is effectively integrated into a pre-trained diffusion model, outperforming existing state-of-the-art methods for panoramic image outpainting. In particular, our SpND method reduces the FID values by more than 50% against the state-of-the-art PanoDiffusion method. Codes are publicly available at https://github.com/chronos123/SpND.
APA
Sun, X., Ma, S., Li, S., Xu, M., Xia, J., Jiang, L., Deng, X. & Wang, J.. (2025). Spherical-Nested Diffusion Model for Panoramic Image Outpainting. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:57614-57630 Available from https://proceedings.mlr.press/v267/sun25m.html.

Related Material