scdrand

Hi, Can you please explain how you collected the BEV semantic image to me? Carla has only 'sensor. camera.semantic_segmentation'. How did you collect the BEV view without postprocessing? Thanks

v2x-setup giving 0 mAP

3

Hi, I used the DeepAccident_tiny_v2x.py setup from the projects/configs as it is. But the mAP score on validation is 0 mAP.

Transformation Matrices

1

Hi, I was trying to transform bboxes from one agent (eg. ego vehicle) to infrastructure using the 'ego_to_world_matrix' in the 'img_metas' or 'relative_ego_pose' in matrix form. But I get the...