We address the problem of unsupervised semantic segmentation of outdoor LiDAR point clouds in diverse traffic scenarios. The key idea is to leverage the spatiotemporal nature of a dynamic point cloud sequence and introduce drastically stronger augmentation by establishing spatiotemporal correspondences across multiple frames. We dovetail clustering and pseudo-label learning in this work. Essentially, we alternate between clustering points into semantic groups and optimizing models using point-wise pseudo-spatiotemporal labels with a simple learning objective. Therefore, our method can learn discriminative features in an unsupervised learning fashion. We show promising segmentation performance on Semantic-KITTI,SemanticPOSS, and FLORIDA benchmark datasets covering scenarios in autonomous vehicle and intersection infrastructure, which is competitive when compared against many existing fully supervised learning methods. This general framework can lead to a unified representation learning approach for LiDAR point clouds incorporating domain knowledge.
A Spatiotemporal Correspondence Approach to Unsupervised LiDAR Segmentation with Traffic Applications
2023-09-24
3830230 byte
Conference paper
Electronic Resource
English
Unsupervised multiphase segmentation: A recursive approach
British Library Online Contents | 2009
|Spatiotemporal Object Database Approach to Dynamic Segmentation
Online Contents | 2003
|Spatiotemporal Object Database Approach to Dynamic Segmentation
Transportation Research Record | 2003
|