SOLiD

[IEEE RA-L] This repository is the official code for Narrowing your FOV with SOLiD: Spatially Organized and Lightweight Global Descriptor for FOV-constrained LiDAR Place Recognition.

Hogyun Kim, Jiwon Choi, Taehu Sim, Giseop Kim, Younggun Cho^†

Spatial AI and Robotics Lab (SPARO)

NEWS

• [May, 2025] SOLiD-SLAM package will be released soon. Stay Tuned!!
• [May, 2025] We will present our paper at ICRA 2025.
• [November, 2024] Now, Distributed-SOLiD-SLAM code is released!!
• [September, 2024] SOLiD is introduced in HeLiPR-Place-Recognition Toolbox!!
• [August, 2024] Now, the SOLiD-A-LOAM code is released!!
• [August, 2024] Now, the SOLiD-PyICP-SLAM code is released!!
• [August, 2024] The SOLiD is added in awesome-lidar-place-recognition!!
• [August, 2024] Now, the SOLiD code is released!!
• [July, 2024] SOLiD is accepted in RA-L!!

TODO

• Integrated with KISS-ICP: SOLiD-KISS-ICP
• Integrated with LIO-SAM: SOLiD-LIO-SAM
• Integrated with FAST-LIO2: SOLiD-FAST-LIO2
• Integrated with LT-mapper

Note

• SOLiD can be integrated with various LiDAR odometry including solid-state LiDAR and Situations (i.e. intra-session, inter-session, and multi-robot).
  • Intra-session SLAM
    • Integrated with A-LOAM: SOLiD-A-LOAM
    • Integrated with a basic ICP odometry: SOLiD-PyICP-SLAM
      • This implementation is fully Python-based so slow but for educational purposes.
  • Inter-session SLAM
    • (TBD) Integrated with LT-mapper
  • Multi-Robot SLAM
    • Integrated with DiSCo SLAM: Distributed-SOLiD-SLAM

What are the problems with traditional LiDAR Place Recognition?

• The traditional method uses a bird eye view and overlooks vertical information.
• Also, because it focuses on performance, it is difficult to apply in real-time on an onboard computer.

  

What is the SOLiD?

• SOLiD is a lightweight and fast LiDAR global descriptor for FOV constraints situations that are limited through fusion with other sensors or blocked by robot/sensor operators including mechanical components or solid-state LiDAR (e.g. Livox).

  

How to use the SOLiD?

• Python version

  • If you use the other LiDAR sensor, you modify the parameters in parser of test.py and lidar file structure in utils/point_module.py.

    $ git clone https://github.com/sparolab/SOLiD.git
    $ cd python
    $ python3 test.py
    

• Cpp version

  • If you use the other LiDAR sensor, you modify the parameters in include/solid_module.h.

      $ git clone https://github.com/sparolab/SOLiD.git
      $ cd cpp
      $ mkdir build
      $ cd build
      $ cmake ..
      $ make
      $ ./test_solid
    

Utils

• If you want to clip the points, you can use python/bin/point_clipper.py

    $ git clone https://github.com/sparolab/SOLiD.git
    $ cd python/bin/
    $ python3 point_clipper.py
  

Supplementary

• Arxiv
• Paper
• Video
• Project page

Main Contribution

• Hogyun Kim
• Jiwon Choi
• Taehu Sim
• Giseop Kim
• Younggun Cho

QnA

• If you have a question, you utilize a alphaXiv and comment here.

Citation

  @article{kim2024narrowing,
    title={Narrowing your FOV with SOLiD: Spatially Organized and Lightweight Global Descriptor for FOV-constrained LiDAR Place Recognition},
    author={Kim, Hogyun and Choi, Jiwon and Sim, Taehu and Kim, Giseop and Cho, Younggun},
    journal={IEEE Robotics and Automation Letters},
    year={2024},
    publisher={IEEE}
  }

Contact

• Hogyun Kim ([email protected])

License

• For academic usage, the code is released under the BSD 3.0 license. For any commercial purpose, please contact the authors.

Contributors