Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2603.16099 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912969654272000 |
|---|---|
| author | Gao, Sensen Wang, Zhaoqing Cao, Qihang Yu, Dongdong Wang, Changhu Liu, Tongliang Gong, Mingming Bian, Jiawang |
| author_facet | Gao, Sensen Wang, Zhaoqing Cao, Qihang Yu, Dongdong Wang, Changhu Liu, Tongliang Gong, Mingming Bian, Jiawang |
| contents | Existing diffusion-based 3D scene generation methods primarily operate in 2D image/video latent spaces, which makes maintaining cross-view appearance and geometric consistency inherently challenging. To bridge this gap, we present OneWorld, a framework that performs diffusion directly within a coherent 3D representation space. Central to our approach is the 3D Unified Representation Autoencoder (3D-URAE); it leverages pretrained 3D foundation models and augments their geometry-centric nature by injecting appearance and distilling semantics into a unified 3D latent space. Furthermore, we introduce token-level Cross-View-Correspondence (CVC) consistency loss to explicitly enforce structural alignment across views, and propose Manifold-Drift Forcing (MDF) to mitigate train-inference exposure bias and shape a robust 3D manifold by mixing drifted and original representations. Comprehensive experiments demonstrate that OneWorld generates high-quality 3D scenes with superior cross-view consistency compared to state-of-the-art 2D-based methods. Our code will be available at https://github.com/SensenGao/OneWorld. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_16099 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | OneWorld: Taming Scene Generation with 3D Unified Representation Autoencoder Gao, Sensen Wang, Zhaoqing Cao, Qihang Yu, Dongdong Wang, Changhu Liu, Tongliang Gong, Mingming Bian, Jiawang Computer Vision and Pattern Recognition Existing diffusion-based 3D scene generation methods primarily operate in 2D image/video latent spaces, which makes maintaining cross-view appearance and geometric consistency inherently challenging. To bridge this gap, we present OneWorld, a framework that performs diffusion directly within a coherent 3D representation space. Central to our approach is the 3D Unified Representation Autoencoder (3D-URAE); it leverages pretrained 3D foundation models and augments their geometry-centric nature by injecting appearance and distilling semantics into a unified 3D latent space. Furthermore, we introduce token-level Cross-View-Correspondence (CVC) consistency loss to explicitly enforce structural alignment across views, and propose Manifold-Drift Forcing (MDF) to mitigate train-inference exposure bias and shape a robust 3D manifold by mixing drifted and original representations. Comprehensive experiments demonstrate that OneWorld generates high-quality 3D scenes with superior cross-view consistency compared to state-of-the-art 2D-based methods. Our code will be available at https://github.com/SensenGao/OneWorld. |
| title | OneWorld: Taming Scene Generation with 3D Unified Representation Autoencoder |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2603.16099 |