Concavity-Induced Distance for Unoriented Point Cloud Decomposition
This addresses the challenge of processing raw point clouds in robotics without meshing or normal estimation, offering a novel approach for instance segmentation and scene representation.
The authors tackled the problem of analyzing unoriented point clouds by proposing Concavity-induced Distance (CID) to measure dissimilarity between points, enabling instance segmentation with comparable average precision to supervised deep learning methods on S3DIS and ScanNet datasets, and improving grouping quality for robotics applications.
We propose Concavity-induced Distance (CID) as a novel way to measure the dissimilarity between a pair of points in an unoriented point cloud. CID indicates the likelihood of two points or two sets of points belonging to different convex parts of an underlying shape represented as a point cloud. After analyzing its properties, we demonstrate how CID can benefit point cloud analysis without the need for meshing or normal estimation, which is beneficial for robotics applications when dealing with raw point cloud observations. By randomly selecting very few points for manual labeling, a CID-based point cloud instance segmentation via label propagation achieves comparable average precision as recent supervised deep learning approaches, on S3DIS and ScanNet datasets. Moreover, CID can be used to group points into approximately convex parts whose convex hulls can be used as compact scene representations in robotics, and it outperforms the baseline method in terms of grouping quality. Our project website is available at: https://ai4ce.github.io/CID/