How do you process data for nuscenes or other LiDAR segmentation dataset?

Pointcept / PointTransformerV3

[CVPR'24 Oral] Official repository of Point Transformer V3 (PTv3)

MIT License

587 stars 30 forks source link

How do you process data for nuscenes or other LiDAR segmentation dataset? #8

Open Akash-Kumbar opened 5 months ago

Akash-Kumbar commented 5 months ago

Considering that those point clouds are sparse, how do you guys process or maybe divide the data into smaller parts and pass it to the network? Are there any resources that I can refer to learn more about the LiDAR point cloud pre-processing.

Thanks in Advance.

Gofinge commented 5 months ago

Hi, maybe you can refer config here (https://github.com/Pointcept/Pointcept/blob/main/configs/nuscenes/semseg-spunet-v1m1-0-base.py) Any questions are welcome. (no need to divide into smaller parts, only 10,000 level points, SpUNet, PTv2, or PTv3 can directly handle them)