MV-SAM3D: Adaptive Multi-View Fusion for Layout-Aware 3D Generation
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914460707323904 |
|---|---|
| author | Li, Baicheng Wu, Dong Li, Jun Zhou, Shunkai Zeng, Zecui Li, Lusong Zha, Hongbin |
| author_facet | Li, Baicheng Wu, Dong Li, Jun Zhou, Shunkai Zeng, Zecui Li, Lusong Zha, Hongbin |
| contents | Recent unified 3D generation models have made remarkable progress in producing high-quality 3D assets from a single image. Notably, layout-aware approaches such as SAM3D can reconstruct multiple objects while preserving their spatial arrangement, opening the door to practical scene-level 3D generation. However, current methods are limited to single-view input and cannot leverage complementary multi-view observations, while independently estimated object poses often lead to physically implausible layouts such as interpenetration and floating artifacts.
We present MV-SAM3D, a training-free framework that extends layout-aware 3D generation with multi-view consistency and physical plausibility. We formulate multi-view fusion as a Multi-Diffusion process in 3D latent space and propose two adaptive weighting strategies -- attention-entropy weighting and visibility weighting -- that enable confidence-aware fusion, ensuring each viewpoint contributes according to its local observation reliability. For multi-object composition, we introduce physics-aware optimization that injects collision and contact constraints both during and after generation, yielding physically plausible object arrangements. Experiments on standard benchmarks and real-world multi-object scenes demonstrate significant improvements in reconstruction fidelity and layout plausibility, all without any additional training. Code is available at https://github.com/devinli123/MV-SAM3D. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_11633 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | MV-SAM3D: Adaptive Multi-View Fusion for Layout-Aware 3D Generation Li, Baicheng Wu, Dong Li, Jun Zhou, Shunkai Zeng, Zecui Li, Lusong Zha, Hongbin Computer Vision and Pattern Recognition Recent unified 3D generation models have made remarkable progress in producing high-quality 3D assets from a single image. Notably, layout-aware approaches such as SAM3D can reconstruct multiple objects while preserving their spatial arrangement, opening the door to practical scene-level 3D generation. However, current methods are limited to single-view input and cannot leverage complementary multi-view observations, while independently estimated object poses often lead to physically implausible layouts such as interpenetration and floating artifacts. We present MV-SAM3D, a training-free framework that extends layout-aware 3D generation with multi-view consistency and physical plausibility. We formulate multi-view fusion as a Multi-Diffusion process in 3D latent space and propose two adaptive weighting strategies -- attention-entropy weighting and visibility weighting -- that enable confidence-aware fusion, ensuring each viewpoint contributes according to its local observation reliability. For multi-object composition, we introduce physics-aware optimization that injects collision and contact constraints both during and after generation, yielding physically plausible object arrangements. Experiments on standard benchmarks and real-world multi-object scenes demonstrate significant improvements in reconstruction fidelity and layout plausibility, all without any additional training. Code is available at https://github.com/devinli123/MV-SAM3D. |
| title | MV-SAM3D: Adaptive Multi-View Fusion for Layout-Aware 3D Generation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2603.11633 |