Abstract: In the field of deep point cloud understanding, KP-Conv is a unique architecture that uses kernel points to locate convolutional weights in space, instead of relying on Multi-Layer ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results