SPCNet: Stepwise Point Cloud Completion Network

TL;DR

SPCNet is a hierarchical, stepwise neural network that progressively completes large-missing 3D point clouds by inferring global, local, and detailed features, inspired by physical object repair.

Contribution

The paper introduces a novel hierarchical, iterative architecture with a cycle loss for robust large-missing point cloud completion, advancing beyond existing methods.

Findings

Outperforms state-of-the-art on large-missing point cloud datasets

Effective in recovering global shape, local details, and fine features

Cycle loss enhances generalization and robustness

Abstract

How will you repair a physical object with large missings? You may first recover its global yet coarse shape and stepwise increase its local details. We are motivated to imitate the above physical repair procedure to address the point cloud completion task. We propose a novel stepwise point cloud completion network (SPCNet) for various 3D models with large missings. SPCNet has a hierarchical bottom-to-up network architecture. It fulfills shape completion in an iterative manner, which 1) first infers the global feature of the coarse result; 2) then infers the local feature with the aid of global feature; and 3) finally infers the detailed result with the help of local feature and coarse result. Beyond the wisdom of simulating the physical repair, we newly design a cycle loss %based training strategy to enhance the generalization and robustness of SPCNet. Extensive experiments clearly show the superiority of our SPCNet over the state-of-the-art methods on 3D point clouds with large missings.