Amodal3R: Amodal 3D Reconstruction from Occluded 2D Images
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866913741103169536 |
|---|---|
| author | Wu, Tianhao Zheng, Chuanxia Guan, Frank Vedaldi, Andrea Cham, Tat-Jen |
| author_facet | Wu, Tianhao Zheng, Chuanxia Guan, Frank Vedaldi, Andrea Cham, Tat-Jen |
| contents | Most image-based 3D object reconstructors assume that objects are fully visible, ignoring occlusions that commonly occur in real-world scenarios. In this paper, we introduce Amodal3R, a conditional 3D generative model designed to reconstruct 3D objects from partial observations. We start from a "foundation" 3D generative model and extend it to recover plausible 3D geometry and appearance from occluded objects. We introduce a mask-weighted multi-head cross-attention mechanism followed by an occlusion-aware attention layer that explicitly leverages occlusion priors to guide the reconstruction process. We demonstrate that, by training solely on synthetic data, Amodal3R learns to recover full 3D objects even in the presence of occlusions in real scenes. It substantially outperforms existing methods that independently perform 2D amodal completion followed by 3D reconstruction, thereby establishing a new benchmark for occlusion-aware 3D reconstruction. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_13439 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Amodal3R: Amodal 3D Reconstruction from Occluded 2D Images Wu, Tianhao Zheng, Chuanxia Guan, Frank Vedaldi, Andrea Cham, Tat-Jen Computer Vision and Pattern Recognition Most image-based 3D object reconstructors assume that objects are fully visible, ignoring occlusions that commonly occur in real-world scenarios. In this paper, we introduce Amodal3R, a conditional 3D generative model designed to reconstruct 3D objects from partial observations. We start from a "foundation" 3D generative model and extend it to recover plausible 3D geometry and appearance from occluded objects. We introduce a mask-weighted multi-head cross-attention mechanism followed by an occlusion-aware attention layer that explicitly leverages occlusion priors to guide the reconstruction process. We demonstrate that, by training solely on synthetic data, Amodal3R learns to recover full 3D objects even in the presence of occlusions in real scenes. It substantially outperforms existing methods that independently perform 2D amodal completion followed by 3D reconstruction, thereby establishing a new benchmark for occlusion-aware 3D reconstruction. |
| title | Amodal3R: Amodal 3D Reconstruction from Occluded 2D Images |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2503.13439 |