MGPC: Multimodal Network for Generalizable Point Cloud Completion With Modality Dropout and Progressive Decoding
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866914282749296640 |
|---|---|
| author | Liu, Jiangyuan Zhao, Yuhao Ma, Hongxuan Liu, Zhe Wang, Jian Zou, Wei |
| author_facet | Liu, Jiangyuan Zhao, Yuhao Ma, Hongxuan Liu, Zhe Wang, Jian Zou, Wei |
| contents | Point cloud completion aims to recover complete 3D geometry from partial observations caused by limited viewpoints and occlusions. Existing learning-based works, including 3D Convolutional Neural Network (CNN)-based, point-based, and Transformer-based methods, have achieved strong performance on synthetic benchmarks. However, due to the limitations of modality, scalability, and generative capacity, their generalization to novel objects and real-world scenarios remains challenging. In this paper, we propose MGPC, a generalizable multimodal point cloud completion framework that integrates point clouds, RGB images, and text within a unified architecture. MGPC introduces an innovative modality dropout strategy, a Transformer-based fusion module, and a novel progressive generator to improve robustness, scalability, and geometric modeling capability. We further develop an automatic data generation pipeline and construct MGPC-1M, a large-scale benchmark with over 1,000 categories and one million training pairs. Extensive experiments on MGPC-1M and in-the-wild data demonstrate that the proposed method consistently outperforms prior baselines and exhibits strong generalization under real-world conditions. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_03660 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | MGPC: Multimodal Network for Generalizable Point Cloud Completion With Modality Dropout and Progressive Decoding Liu, Jiangyuan Zhao, Yuhao Ma, Hongxuan Liu, Zhe Wang, Jian Zou, Wei Computer Vision and Pattern Recognition Point cloud completion aims to recover complete 3D geometry from partial observations caused by limited viewpoints and occlusions. Existing learning-based works, including 3D Convolutional Neural Network (CNN)-based, point-based, and Transformer-based methods, have achieved strong performance on synthetic benchmarks. However, due to the limitations of modality, scalability, and generative capacity, their generalization to novel objects and real-world scenarios remains challenging. In this paper, we propose MGPC, a generalizable multimodal point cloud completion framework that integrates point clouds, RGB images, and text within a unified architecture. MGPC introduces an innovative modality dropout strategy, a Transformer-based fusion module, and a novel progressive generator to improve robustness, scalability, and geometric modeling capability. We further develop an automatic data generation pipeline and construct MGPC-1M, a large-scale benchmark with over 1,000 categories and one million training pairs. Extensive experiments on MGPC-1M and in-the-wild data demonstrate that the proposed method consistently outperforms prior baselines and exhibits strong generalization under real-world conditions. |
| title | MGPC: Multimodal Network for Generalizable Point Cloud Completion With Modality Dropout and Progressive Decoding |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2601.03660 |