Search papers, labs, and topics across Lattice.
MV-SAM3D is introduced as a training-free framework that extends layout-aware 3D generation to leverage multi-view consistency and physical plausibility. It formulates multi-view fusion as a Multi-Diffusion process in 3D latent space, using attention-entropy and visibility weighting for confidence-aware fusion. Physics-aware optimization is employed to enforce collision and contact constraints during and after generation, improving layout plausibility.
Achieve significantly improved 3D reconstruction fidelity and layout plausibility in multi-object scenes by fusing multi-view observations with physics-aware constraints, all without requiring additional training.
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.
