SegGroup: Seg-Level Supervision for 3D Instance and Semantic Segmentation
This work tackles the problem of high annotation costs for 3D point cloud instance and semantic segmentation, which is a significant practical challenge for researchers and practitioners working with 3D data.
This paper addresses the high annotation cost of 3D point cloud segmentation by proposing a weakly-supervised method that only requires clicking one point per instance for annotation. The method, SegGroup, extends these single-point annotations to segment-level labels and then uses a network to hierarchically group unlabeled segments, generating point-level pseudo labels. This approach achieves results comparable to fully-supervised methods and outperforms other weakly-supervised methods under the same annotation budget.
Most existing point cloud instance and semantic segmentation methods rely heavily on strong supervision signals, which require point-level labels for every point in the scene. However, such strong supervision suffers from large annotation costs, arousing the need to study efficient annotating. In this paper, we discover that the locations of instances matter for both instance and semantic 3D scene segmentation. By fully taking advantage of locations, we design a weakly-supervised point cloud segmentation method that only requires clicking on one point per instance to indicate its location for annotation. With over-segmentation for pre-processing, we extend these location annotations into segments as seg-level labels. We further design a segment grouping network (SegGroup) to generate point-level pseudo labels under seg-level labels by hierarchically grouping the unlabeled segments into the relevant nearby labeled segments, so that existing point-level supervised segmentation models can directly consume these pseudo labels for training. Experimental results show that our seg-level supervised method (SegGroup) achieves comparable results with the fully annotated point-level supervised methods. Moreover, it outperforms the recent weakly-supervised methods given a fixed annotation budget. Code is available at https://github.com/AnTao97/SegGroup.