RE-PU: A Self-Supervised Arbitrary-Scale Point Cloud Upsampling Method Based on Reconstruction
Open Access
- 5 August 2024
- journal article
- research article
- Published by MDPI AG in Applied Sciences
- Vol. 14 (15) , 6814
Abstract
The point clouds obtained directly from three-dimensional scanning devices are often sparse and noisy. Therefore, point cloud upsampling plays an increasingly crucial role in fields such as point cloud reconstruction and rendering. However, point cloud upsampling methods are primarily supervised and fixed-rate, which restricts their applicability in various scenarios. In this paper, we propose a novel point cloud upsampling method, named RE-PU, which is based on the point cloud reconstruction and achieves self-supervised upsampling at arbitrary rates. The proposed method consists of two main stages: the first stage is to train a network to reconstruct the original point cloud from a prior distribution, and the second stage is to upsample the point cloud data by increasing the number of sampled points on the prior distribution with the trained model. The experimental results demonstrate that the proposed method can achieve comparable outcomes to supervised methods in terms of both visual quality and quantitative metrics.Keywords
Funding Information
- National Natural Science Foundation of China (62362004)
This publication has 48 references indexed in Scilit:
- PU-Ray: Domain-Independent Point Cloud Upsampling via Ray Marching on Neural Implicit SurfaceIEEE Transactions on Intelligent Transportation Systems, 2024
- Unsupervised Point Cloud Representation Learning With Deep Neural Networks: A SurveyIEEE Transactions on Pattern Analysis and Machine Intelligence, 2023
- PUFA-GAN: A Frequency-Aware Generative Adversarial Network for 3D Point Cloud UpsamplingIEEE Transactions on Image Processing, 2022
- PU-Flow: A Point Cloud Upsampling Network With Normalizing FlowsIEEE Transactions on Visualization and Computer Graphics, 2022
- PU-Dense: Sparse Tensor-Based Point Cloud Geometry UpsamplingIEEE Transactions on Image Processing, 2022
- SPU-Net: Self-Supervised Point Cloud Upsampling by Coarse-to-Fine Reconstruction With Self-Projection OptimizationIEEE Transactions on Image Processing, 2022
- Deep Magnification-Flexible Upsampling Over 3D Point CloudsIEEE Transactions on Image Processing, 2021
- PCT: Point cloud transformerComputational Visual Media, 2021
- Meta-PU: An Arbitrary-Scale Upsampling Network for Point CloudIEEE Transactions on Visualization and Computer Graphics, 2021
- Deep Unsupervised Learning of 3D Point Clouds via Graph Topology Inference and FilteringIEEE Transactions on Image Processing, 2019