SG-SLAM

Leveraging Semantic Graphs for Efficient and Robust LiDAR SLAM

Neng Wang1, Huimin Lu1, Zhiqiang Zheng1, Hesheng Wang2, Yun-Hui Liu3 Xieyuanli Chen1
1National University of Defense Technology, 2Shanghai Jiao Tong University, 3Chinese University of Hong Kong
Paper Code
SG-SLAM picture

Abstract

we propose a semantic graph-enhanced SLAM framework, named SG-SLAM, which effectively leverages the geometric, semantic, and topological characteristics inherent in environmental structures.

The semantic graph serves as a fundamental component that facilitates critical functionalities of SLAM, including robust relocalization during odometry failures, accurate loop closing, and semantic graph map construction.

Our method employs a dual-threaded architecture, with one thread dedicated to online odometry and relocalization, while the other handles loop closure, pose graph optimization, and map update.

This design enables our method to operate in real time and generate globally consistent semantic graph maps and point cloud maps. We extensively evaluate our method across the KITTI, MulRAN, and Apollo datasets, and the results demonstrate its superiority compared to state-of-the-art methods.

SLAM Demo

SLAM Demo

Odometry with Graph-based Relocalization


Demo

This demo illustrates that our semantic graph-based relocalization algorithm significantly enhances the robustness of the odometry.

SLAM Demo

Path

We compare the path of odometry using SG-SLAM with the semantic graph-based relocalization (green) and SG-SLAM without the relocalization (blue).

SLAM Demo

BibTeX


      @article{wang2025sgslam,
        title     = {{Leveraging Semantic Graphs for Efficient and Robust LiDAR SLAM}},
        author    = {Neng, Wang and Huimin, Lu and Zhiqiang, Zheng and Hesheng, Wang and Yun-Hui, Liu and Xieyuanli, Chen},
        journal   = {arXiv preprint arXiv:2503.11145},
        year      = {2025}
      }