PosePN

LiDAR-based localization using universal encoding and memory-aware regression

Environment

• python

• pytorch

• MinkowskiEngine

Data

We support the Oxford Radar RobotCar, vReLoc, and NCLT datasets right now.

Oxford data_root
├── 2019-01-11-14-02-26-radar-oxford-10k
│   ├── velodyne_left
│       ├── xxx.bin
├── pose_stats.txt
├── pose_max_min.txt
├── train_split.txt
├── test_split.txt

Run

Oxford, vReLoc, NCLT

• train -- 1 GPU

python train.py

• test -- 1 GPU

python eval.py

Acknowledgement

We appreciate the code of PointNet, PointNet++, SOE-Net, MinkLoc3D, and AtLoc they shared.

Citation

@article{YU2022108685,
title = {LiDAR-based localization using universal encoding and memory-aware regression},
journal = {Pattern Recognition},
volume = {128},
pages = {108685},
year = {2022},
issn = {0031-3203},
doi = {https://doi.org/10.1016/j.patcog.2022.108685},
author = {Shangshu Yu and Cheng Wang and Chenglu Wen and Ming Cheng and Minghao Liu and Zhihong Zhang and Xin Li}
}