search

PRBonn / lidar-bonnetal

Semantic and Instance Segmentation of LiDAR point clouds for autonomous driving
http://semantic-kitti.org
MIT License
961 stars 206 forks source link

about the dataest #114

Closed zzhmx closed 3 months ago

zzhmx commented 8 months ago

May I ask how many lines of point cloud data are scanned in your dataset? After converting my 128 line radar into range images, it is still very sparse. If there are too many clicks, an error will be reported: Runtimeerror the expanded size of the tensor (150000) must match the existing size (159400) at non-singleton dimension 0. Target size (150000,3). Tensor size:(159400,3)

jbehley commented 7 months ago

I think the 150k points is just an upper bound that we used for the KITTI dataset (with 64 beams). Therefore, you might have to simply increase this in the configuration, e.g., https://github.com/PRBonn/lidar-bonnetal/blob/master/train/tasks/semantic/config/arch/darknet53.yaml

https://github.com/PRBonn/lidar-bonnetal/blob/5a5f4b180117b08879ec97a3a05a3838bce6bb0f/train/tasks/semantic/config/arch/darknet53.yaml#L77-L100

zzhmx commented 7 months ago

I think the 150k points is just an upper bound that we used for the KITTI dataset (with 64 beams). Therefore, you might have to simply increase this in the configuration, e.g., https://github.com/PRBonn/lidar-bonnetal/blob/master/train/tasks/semantic/config/arch/darknet53.yaml我认为 150k 个点只是我们用于 KITTI 数据集（有 64 个光束）的上限。因此，您可能只需在配置中增加此设置，例如 https://github.com/PRBonn/lidar-bonnetal/blob/master/train/tasks/semantic/config/arch/darknet53.yaml

https://github.com/PRBonn/lidar-bonnetal/blob/5a5f4b180117b08879ec97a3a05a3838bce6bb0f/train/tasks/semantic/config/arch/darknet53.yaml#L77-L100 thank you!