Fetching the paper…

CAPE: Camera View Position Embedding for Multi-View 3D Object Detection · Around