SEPT: Towards Efficient Scene Representation Learning for Motion Prediction

Open in new window