Fetching the paper…

AutoAlignV2: Deformable Feature Aggregation for Dynamic Multi-Modal 3D Object Detection · Around