Marginally calibrated response distributions for end-to-end learning in autonomous driving

Open in new window