Introduced in View-Guided Point Cloud Completion
A large-scale dataset for the point cloud completion task on the ShapeNet dataset.
This paper addresses 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.
