In one frame of the KITTI dataset,

[abo1arab]

In one frame of the KITTI dataset, I found a truck clearly visible in the left camera image and labeled in the annotation file, but completely missing from the LiDAR point cloud, making it impossible to annotate in 3D — this might be due to LiDAR occlusion, angle limitations, or poor reflection from the truck’s surface; should I ignore such objects during 3D annotation, or is there a recommended approach?"