Multi-scale Interaction for Real-time LiDAR Data Segmentation on an Embedded Platform (RA-L)
Dependence:
- Accoding to LiDAR-Bonnetal (https://github.com/PRBonn/lidar-bonnetal/tree/master/train)
- flops-counter.pytorch (https://github.com/sovrasov/flops-counter.pytorch)
- Edge files: Edges
Infer:
- put 'sequences' folder under 'data/'
- 'python infer.py --dataset data --arch_cfg config/arch/config_file --data_cfg config/labels/semantic-kitti.yaml --checkpoint checkpoints/checkpoint_file --log predictions'
Attention
- Only infer validation set, refer 'lib/user.py' line 70-80.
- pay attention to kNN setting. (In 'config/arch/*.yaml')
Citation
Please cite the following paper if you use this repository in your reseach.
@ARTICLE{9633188,
author={Li, Shijie and Chen, Xieyuanli and Liu, Yun and Dai, Dengxin and Stachniss, Cyrill and Gall, Juergen},
journal={IEEE Robotics and Automation Letters},
title={Multi-Scale Interaction for Real-Time LiDAR Data Segmentation on an Embedded Platform},
year={2022},
volume={7},
number={2},
pages={738-745},
doi={10.1109/LRA.2021.3132059}}