Abstract
3D scene understanding, e.g., point cloud semantic and instance segmentation, often requires large-scale annotated training data, but clearly, point-wise labels are too tedious to prepare. While some recent methods propose to train a 3D network with small percentages of point labels, we take the approach to an extreme and propose “One Thing One Click,” meaning that the annotator only needs to label one point per object. To leverage these extremely sparse labels in network training, we design a novel self-training approach, in which we iteratively conduct the training and label propagation, facilitated by a graph propagation module. Also, we adopt a relation network to generate the per-category prototype to enhance the pseudo label quality and guide the iterative training. Besides, our model can be compatible to 3D instance segmentation equipped with a point-clustering strategy. Experimental results on both ScanNet-v2 and S3DIS show that our self-training approach, with extremely sparse annotations, outperforms all existing weakly supervised methods for 3D semantic and instance segmentation by a large margin, and our results are also comparable to those of the fully supervised counterparts. Codes and models are available at https://github.com/liuzhengzhe/One-Thing-One-Click.
Original language | English |
---|---|
Title of host publication | Deep Learning for 3D Vision: Algorithms and Applications |
Editors | Xiaoli LI, Xulei YANG, Hao SU |
Publisher | World Scientific Publishing Co. |
Chapter | 3 |
Pages | 57-89 |
Number of pages | 33 |
ISBN (Electronic) | 9789811286490, 9789811286506 |
ISBN (Print) | 9789811286483 |
DOIs | |
Publication status | Published - Sept 2024 |
Externally published | Yes |
Bibliographical note
Publisher Copyright:© 2024 by World Scientific Publishing Co. Pte. Ltd.