MILAN: Milli-Annotations for Lidar Semantic Segmentation
Preprint
Abstract
Annotating lidar point clouds for autonomous driving is a notoriously expensive and time-consuming task. In this work, we show that the quality of recent self-supervised lidar scan representations allows a great reduction of the annotation cost. Our method has two main steps. First, we show that self-supervised representations allow a simple and direct selection of highly informative lidar scans to annotate: training a network on these selected scans leads to much better results than a random selection of scans and, more interestingly, to results on par with selections made by SOTA active learning methods. In a second step, we leverage the same self-supervised representations to cluster points in our selected scans. Asking the annotator to classify each cluster, with a single click per cluster, then permits us to close the gap with fully-annotated training sets, while only requiring one thousandth of the point labels.Keywords
All Related Versions
This publication has 7 references indexed in Scilit:
- BaSAL: Size-Balanced Warm Start Active Learning for LiDAR Semantic Segmentation2024
- Weakly supervised semantic segmentation of mobile laser scanning point clouds via category balanced random annotation and deep consistency‐guided self‐distillation mechanismThe Photogrammetric Record, 2023
- Using a Waffle Iron for Automotive Point Cloud Semantic Segmentation2023
- Discwise Active Learning for LiDAR Semantic SegmentationIEEE Robotics and Automation Letters, 2023
- Supervoxel-based and Cost-Effective Active Learning for Point Cloud Semantic Segmentation2022
- Exploiting Diversity of Unlabeled Data for Label-Efficient Semi-Supervised Active Learning2022
- SegContrast: 3D Point Cloud Feature Representation Learning Through Self-Supervised Segment DiscriminationIEEE Robotics and Automation Letters, 2022