Unsupervised Point Cloud Pre-Training via View-Point Occlusion Completion

Publication
In International Conference on Computer Vision