PETR: Position Embedding Transformation for Multi-View 3D Object Detection

In this paper, we develop position embedding transformation (PETR) for\nmulti-view 3D object detection. PETR encodes the position information of 3D\ncoordinates into image features, producing the 3D position-aware features.\nObject query can perceive the 3D position-aware features and perform end-to-end\nobject detection. PETR achieves state-of-the-art performance (50.4% NDS and\n44.1% mAP) on standard nuScenes dataset and ranks 1st place on the benchmark.\nIt can serve as a simple yet strong baseline for future research. Code is\navailable at \\url{https://github.com/megvii-research/PETR}.\n