Open zjr-bit opened 8 months ago
Hi! Thanks for your great work. I have been reading your code about computing metrics of planning, and I have a question. As you mentioned in https://github.com/OpenDriveLab/UniAD/issues/20#issue-1691603823, the predicted waypoints are in the lidar frame, which means upwards and to the right are considered positive. But in BEV frame, the upper-left corner is (0,0), and the lower-right corner is (199,199). So based on the above, it seems that the below code snippet conflicts with the frame definition. https://github.com/OpenDriveLab/UniAD/blob/ffac1a69a6d01c9dca5cd2ed751007896df351f8/projects/mmdet3d_plugin/uniad/dense_heads/planning_head_plugin/planning_metrics.py#L63-L64 Have I overlooked some critical details? Looking forward to your anwser! Thanks!
Hi! Thanks for your great work. I have been reading your code about computing metrics of planning, and I have a question. As you mentioned in https://github.com/OpenDriveLab/UniAD/issues/20#issue-1691603823, the predicted waypoints are in the lidar frame, which means upwards and to the right are considered positive. But in BEV frame, the upper-left corner is (0,0), and the lower-right corner is (199,199). So based on the above, it seems that the below code snippet conflicts with the frame definition. https://github.com/OpenDriveLab/UniAD/blob/ffac1a69a6d01c9dca5cd2ed751007896df351f8/projects/mmdet3d_plugin/uniad/dense_heads/planning_head_plugin/planning_metrics.py#L63-L64 Have I overlooked some critical details? Looking forward to your anwser! Thanks!