LAWA: LiDAR Adverse Weather Augmentation for Robust Point Cloud Semantic Segmentation

LAWA: LiDAR Adverse Weather Augmentation for Robust Point Cloud Semantic Segmentation

The code will be uploaded after the review process!

Demo

Qualitative comparison of scatter points based on precipitation levels

Comparison with:

• [LISA]: Kilic, Velat, et al. "Lidar light scattering augmentation (lisa): Physics-based simulation of adverse weather conditions for 3d object detection." arXiv preprint arXiv:2107.07004 (2021).
• [Snow-sim]: Hahner, Martin, et al. "Lidar snowfall simulation for robust 3d object detection." Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2022.

<img src="./images/figure_scatter_result.png" width="800">

Result of semantic segmentation using real world adverse weather dataset

• KONKUK Ailab dataset

  <img src="./images/figure_real_quality_result.png" width="800">

• KITTI Dataset

  <img src="./images/quality_result.png" width="800">

Contact

If you have any questions, please let me know:

• Jonghyun Lee (mickey9624@gmail.com)
• Hyunwook Kang (pd3518@gmail.com)