SGTD

SGTD: A Semantic-Guided Triangular Descriptor for One-Shot LiDAR-Based Global Localization

Our code will be made public after the paper is accepted.

Demonstration

Watch our demonstration video on YouTube:

Installation

Prerequisites

Install required dependencies:

sudo apt-get install nlohmann-json3-dev
sudo apt-get install libeigen3-dev
sudo apt install libgtsam-dev libgtsam-unstable-dev

Note: Please install ceres-solver (version > 2.1.0) by following the guide at ceres Installation.

Setup & Usage

# Clone the repository
git clone https://github.com/Hfx-J/SGTD.git

# Navigate to project directory
cd SGTD 

# Build the project
catkin_make -DPYTHON_EXECUTABLE=/usr/bin/python3

# Setup environment
source ./devel/setup.bash

# Launch the application
roslaunch sgtd semantic_graph_localization.launch

DataSet

We provide an example dataset in GoogleDrive. After downloading, you can extract it to the data/ directory and modify the data_process/dataset in src/sgtd/config/SG_localization.yaml accordingly (note: absolute path is required).

Citation

If you find our work useful in your research, please consider citing:

@article{huang2025sgtd,
  title={SGTD: A Semantic-Guided Triangle Descriptor for One-Shot LiDAR-Based Global Localization},
  author={Huang, Feixuan and Gao, Wang and Pan, Shuguo and Liu, Hong and Zhao, Heng},
  journal={IEEE Robotics and Automation Letters},
  year={2025},
  publisher={IEEE}
}

Contact

For any questions or issues regarding SGTD, please open an issue on GitHub.