PU-Net: Point Cloud Upsampling Network

TL;DR

PU-Net introduces a deep learning approach for upsampling 3D point clouds, effectively increasing point density while maintaining surface accuracy and uniform distribution, outperforming existing methods.

Contribution

The paper proposes a novel multi-branch convolutional network that learns multi-level features for point cloud upsampling at the patch level, improving surface adherence and point uniformity.

Findings

Outperforms baseline and optimization-based methods.

Produces more uniform point distributions.

Points are closer to the underlying surface.

Abstract

Learning and analyzing 3D point clouds with deep networks is challenging due to the sparseness and irregularity of the data. In this paper, we present a data-driven point cloud upsampling technique. The key idea is to learn multi-level features per point and expand the point set via a multi-branch convolution unit implicitly in feature space. The expanded feature is then split to a multitude of features, which are then reconstructed to an upsampled point set. Our network is applied at a patch-level, with a joint loss function that encourages the upsampled points to remain on the underlying surface with a uniform distribution. We conduct various experiments using synthesis and scan data to evaluate our method and demonstrate its superiority over some baseline methods and an optimization-based method. Results show that our upsampled points have better uniformity and are located closer to the underlying surfaces.