Closed Jaywxy closed 1 month ago
Hi @Jaywxy ,
For point cloud semseg, you'll need lidar_calibration
, lidar_segmentation
, lidar_pose
, lidar
, vehicle_pose
minimally. You can convert LiDAR range image to 3D point cloud using this function in v2/perception/utils/lidar_utils.py and convert the corresponding segmentation labels as in this answer.
Hi @Jaywxy ,
For point cloud semseg, you'll need
lidar_calibration
,lidar_segmentation
,lidar_pose
,lidar
,vehicle_pose
minimally. You can convert LiDAR range image to 3D point cloud using this function in v2/perception/utils/lidar_utils.py and convert the corresponding segmentation labels as in this answer.
thank you for your reply!
My task is point cloud semantic segmentation, I only need point cloud data, I want to use the V2.0.1 dataset, I get all the data named with lidar, do I need to download other data?Another problem is that the tutorial_3d_semseg.ipynb in the tutorial doesn't seem to work, because the prompt needs a tfrecord file, but I found that the 2.0.1 version of the data does not have this file, and all are files of another formatWhat should I do? Hope to get some advice, thanks