View-Guided Point Cloud Completion
About
This paper presents a view-guided solution for the task of point cloud completion. Unlike most existing methods directly inferring the missing points using shape priors, we address this task by introducing ViPC (view-guided point cloud completion) that takes the missing crucial global structure information from an extra single-view image. By leveraging a framework that sequentially performs effective cross-modality and cross-level fusions, our method achieves significantly superior results over typical existing solutions on a new large-scale dataset we collect for the view-guided point cloud completion task.
Xuancheng Zhang, Yutong Feng, Siqi Li, Changqing Zou, Hai Wan, Xibin Zhao, Yandong Guo, Yue Gao• 2021
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Point Cloud Completion | ShapeNet-ViPC supervised (test) | Mean F-Score @ 0.001 (Avg)59.1 | 16 | |
| Point Cloud Completion | ShapeNet-ViPC (known categories) | Avg Score0.591 | 13 | |
| Point Cloud Completion | ShapeNet-ViPC (test) | MCD (Airplane)1.76 | 10 | |
| Point Cloud Completion | ShapeNet-ViPC Novel categories | MCD (Avg)4.601 | 8 | |
| Point Cloud Completion | ShapeNet-ViPC Novel categories (test) | F-Score@0.001 (Avg)49.8 | 8 | |
| Point Cloud Completion | ShapNet-ViPC | mIoU (Avg)59.1 | 5 |
Showing 6 of 6 rows