Point Cloud Completion Via Skeleton-Detail Transformer

Point cloud shape completion plays a central role in diverse 3D vision and robotics applications. Early methods used to generate global shapes without local detail refinement. Current methods tend to leverage local features to preserve the observed geometric details. However, they usually adopt the convolutional architecture over the incomplete point cloud to extract local features to restore the diverse information of both latent shape skeleton and geometric details, where long-distance correlation among the skeleton and details is ignored. In this work, we present a coarse-to-fine completion framework, which makes full use of both neighboring and long-distance region cues for point cloud completion. Our network leverages a Skeleton-Detail Transformer, which contains cross-attention and self-attention layers, to fully explore the correlation from local patterns to global shape and utilize it to enhance the overall skeleton. Also, we propose a selective attention mechanism to save memory usage in the attention process without significantly affecting performance. We conduct extensive experiments on the ShapeNet dataset and real-scanned datasets. Qualitative and quantitative evaluations demonstrate that our proposed network outperforms current state-of-the-art methods.

Automated summary

Point cloud shape completion is important in 3D vision and robotics applications. Early methods generated global shapes without refining local details. Current methods use local features to preserve observed geometric details, but they ignore long-distance correlation between skeleton and details. In this work, we propose a coarse-to-fine completion framework that leverages neighboring and long-distance cues, and introduces a Skeleton-Detail Transformer and selective attention mechanism. Experimental results show that our network outperforms state-of-the-art methods.