Yang Chen

We employ contrastive learning to extract complete point cloud structures from partial (incomplete) point clouds for guiding point cloud completion, achieving state-of-the-art (SOTA) results in the field of self-supervised point cloud completion.