Voint Cloud: Multi-View Point Cloud Representation for 3D Understanding
This addresses the challenge of integrating multi-view and point cloud data for 3D classification, shape retrieval, and part segmentation, with incremental improvements over existing methods.
The paper tackles the problem of combining multi-view projection methods with 3D point clouds for 3D understanding tasks by introducing a multi-view point cloud (Voint cloud) representation, achieving state-of-the-art performance on benchmarks like ScanObjectNN, ShapeNet Core55, and ShapeNet Parts.
Multi-view projection methods have demonstrated promising performance on 3D understanding tasks like 3D classification and segmentation. However, it remains unclear how to combine such multi-view methods with the widely available 3D point clouds. Previous methods use unlearned heuristics to combine features at the point level. To this end, we introduce the concept of the multi-view point cloud (Voint cloud), representing each 3D point as a set of features extracted from several view-points. This novel 3D Voint cloud representation combines the compactness of 3D point cloud representation with the natural view-awareness of multi-view representation. Naturally, we can equip this new representation with convolutional and pooling operations. We deploy a Voint neural network (VointNet) to learn representations in the Voint space. Our novel representation achieves \sota performance on 3D classification, shape retrieval, and robust 3D part segmentation on standard benchmarks ( ScanObjectNN, ShapeNet Core55, and ShapeNet Parts).