The instance segmentation task in computer vision involves labeling each pixel in an image with a class and an instance label. It can be thought of as a generalization of the semantic segmentation ...

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 ...