Semi-supervised 3D shape segmentation with multilevel consistency and part substitution
This addresses the data scarcity issue in 3D segmentation for computer vision applications, but it is incremental as it builds on semi-supervised techniques.
The paper tackles the problem of limited fine-grained 3D shape segmentation data by proposing a semi-supervised method that uses a multilevel consistency loss for unlabeled data and part substitution for labeled data, achieving superior performance on PartNet, ShapeNetPart, and ScanNet benchmarks compared to existing approaches.
The lack of fine-grained 3D shape segmentation data is the main obstacle to developing learning-based 3D segmentation techniques. We propose an effective semi-supervised method for learning 3D segmentations from a few labeled 3D shapes and a large amount of unlabeled 3D data. For the unlabeled data, we present a novel multilevel consistency loss to enforce consistency of network predictions between perturbed copies of a 3D shape at multiple levels: point-level, part-level, and hierarchical level. For the labeled data, we develop a simple yet effective part substitution scheme to augment the labeled 3D shapes with more structural variations to enhance training. Our method has been extensively validated on the task of 3D object semantic segmentation on PartNet and ShapeNetPart, and indoor scene semantic segmentation on ScanNet. It exhibits superior performance to existing semi-supervised and unsupervised pre-training 3D approaches. Our code and trained models are publicly available at https://github.com/isunchy/semi_supervised_3d_segmentation.