Several computer vision tasks require perceiving or interacting with 3D environments and objects therein, making a strong case in favor of 3D deep learning. However, unlike images which are most ...