Fig. 5From: CenterTransFuser: radar point cloud and visual information fusion for 3D object detectionThe cross-modal attention model consisting of two branches, with the query matrix of the radar branch guiding the image for cross-modal interactionBack to article page