pk1996

Sure I can give it a try!

Hi Thanks for the reply! I will go through your codebase. My changes are here - https://github.com/open-mmlab/OpenPCDet/compare/master...pk1996:OpenPCDet:master 1. I did ensure that the coordinate system of my data matches that...

This is confusing! The fact that the authors use [rect_to_lidar](https://github.com/open-mmlab/OpenPCDet/blob/c233477a4a4bf2861a978794d6b9c047f64641f7/pcdet/utils/calibration_kitti.py#L50) to translate the 3D location to the Velodyne frame of reference [here](https://github.com/open-mmlab/OpenPCDet/blob/c233477a4a4bf2861a978794d6b9c047f64641f7/pcdet/datasets/kitti/kitti_dataset.py#L194), we can tell that the kitti velodyne/lidar frame...

Glad my interpretation of the coordinate system is right! Hope the repo maintainers can shed some light on the angle-related doubt! BTW @OrangeSodahub did you consider modifying the anchor boxes...

https://terra-1-g.djicdn.com/65c028cd298f4669a7f0e40e50ba1131/Download/Avia/readme_EN.md refer to this.

Try this. https://github.com/strawlab/python-pcl/pull/371