# follow uniad, add a 0.5m offset
ego_box[0] += 0.5 * torch.cos(ego_box[6])
ego_box[1] += 0.5 * torch.sin(ego_box[6])
planning_eval.py
In L 24, the comments say: ego_box: tensor with shape [7], [x, y, z, w, l, h, yaw].
These codes are not adding a safety distance surrounding the ego-vehicle like in UniAD:
self.w = 1.85 + delta
self.h = 4.084 + delta
It instead translates the ego-car center by 0.5m in the direction of yaw. I don't see how this could be useful for checking collisions.
I could not find such a translation in the evaluation either.
https://github.com/OpenDriveLab/UniAD/blob/v2.0/projects/mmdet3d_plugin/uniad/dense_heads/planning_head_plugin/planning_metrics.py
Hope you can clarify. Thank you.
planning_eval.py
In L 24, the comments say:
ego_box: tensor with shape [7], [x, y, z, w, l, h, yaw].These codes are not adding a safety distance surrounding the ego-vehicle like in UniAD:
It instead translates the ego-car center by 0.5m in the direction of yaw. I don't see how this could be useful for checking collisions.
I could not find such a translation in the evaluation either.
https://github.com/OpenDriveLab/UniAD/blob/v2.0/projects/mmdet3d_plugin/uniad/dense_heads/planning_head_plugin/planning_metrics.py
Hope you can clarify. Thank you.