PU-Net: Point Cloud Upsampling Network

2018/01/21 by Lequan Yu, Yu, Lequan, Xianzhi Li +7 · 19 citations
Computer Science · Earth and Planetary Sciences · Engineering · #3D Shape Modeling and Analysis #3D Surveying and Cultural Heritage #Computer Graphics and Visualization Techniques #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Graphics (cs.GR)

paper · pdf · doi:10.48550/arxiv.1801.06761

openalex publication_date 2018/01/21 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

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.

Citations

• Adam: A Method for Stochastic Optimization
• Consolidation of unorganized point clouds for surface reconstruction
• ImageNet classification with deep convolutional neural networks
• ShapeNet: An Information-Rich 3D Model Repository
• PointNet: Deep Learning on Point Sets for 3D Classification and\n Segmentation
• Shape Completion using 3D-Encoder-Predictor CNNs and Shape Synthesis
• PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space
• Fully Convolutional Networks for Semantic Segmentation
• ShapeNet: An Information-Rich 3D Model Repository

Cited by

• PUFM++: Point Cloud Upsampling via Enhanced Flow Matching
• Secure AI-Driven Super-Resolution for Real-Time Mixed Reality Applications
• SurfFill: Completion of LiDAR Point Clouds via Gaussian Surfel Splatting
• SplatSuRe: Selective Super-Resolution for Multi-view Consistent 3D Gaussian Splatting
• IPCD: Intrinsic Point-Cloud Decomposition
• 2D Representation for Unguided Single-View 3D Super-Resolution in Real-Time
• 3D-ANC: Adaptive Neural Collapse for Robust 3D Point Cloud Recognition
• U-CAN: Unsupervised Point Cloud Denoising with Consistency-Aware Noise2Noise Matching
• UGAE: Unified Geometry and Attribute Enhancement for G-PCC Compressed Point Clouds
• PointNSP: Autoregressive 3D Point Cloud Generation with Next-Scale Level-of-Detail Prediction
• PU-Gaussian: Point Cloud Upsampling using 3D Gaussian Representation
• Adaptive and Iterative Point Cloud Denoising with Score-Based Diffusion Model
• PVNet: Point-Voxel Interaction LiDAR Scene Upsampling Via Diffusion Models
• A Survey of Deep Learning-based Point Cloud Denoising
• PanoLAM: Large Avatar Model for Gaussian Full-Head Synthesis from One-shot Unposed Image
• Point upsampling networks for single-photon sensing
• Self-Supervised Temporal Super-Resolution of Energy Data using Generative Adversarial Transformer
• Hybrid Long and Short Range Flows for Point Cloud Filtering
• Rate-distortion Optimized Point Cloud Preprocessing for Geometry-based Point Cloud Compression

Related

• PU-GCN: Point Cloud Upsampling using Graph Convolutional Networks
• Meta-PU: An Arbitrary-Scale Upsampling Network for Point Cloud
• Point Cloud Upsampling via Cascaded Refinement Network
• SPU-Net: Self-Supervised Point Cloud Upsampling by Coarse-to-Fine Reconstruction with Self-Projection Optimization
• Arbitrary-Scale Point Cloud Upsampling by Voxel-Based Network with Latent Geometric-Consistent Learning