ROS Relative Learning
目录
Chen, Xieyuanli, Thomas Läbe, Lorenzo Nardi, Jens Behley, and Cyrill Stachniss. “Learning an Overlap-based Observation Model for 3D LiDAR Localization.”
Paper: Overlap-based
Summary
- 文章使用了OverlapNet作为蒙特卡洛定位算法(MCL)的观测模型,实现了基于激光雷达传感器的高精度全局定位。目前MCL最大的难题就是如何去设计一个好的观测模型。文章的创新点是利用OverlapNet来训练了一个观测模型,然后把它集成到MCL中,提高了MCL的定位性能。
- a approach for global localization using 3D Lidar scans on road vehicles;
- novel observation model that exploit the overlap and yaw angle estimation;
- using overlapNet2020 model;
- 开源代码:https://github.com/PRBonn/overlap_localization