DualGenerator: Information Interaction-based Generative Network for Point Cloud Completion
Authors:
- Pengcheng ShiPSPengcheng Shi
- Haozhe ChengHCHaozhe Cheng
- Xu HanXHXu Han
- Yiyang ZhouYZYiyang Zhou
- Jihua ZhuJihua Zhu
Preprint
Abstract
Point cloud completion estimates complete shapes from incomplete point clouds to obtain higher-quality point cloud data. Most existing methods only consider global object features, ignoring spatial and semantic information of adjacent points. They cannot distinguish structural information well between different object parts, and the robustness of models is poor. To tackle these challenges, we propose an information interaction-based generative network for point cloud completion ($\mathbf{DualGenerator}$). It contains an adversarial generation path and a variational generation path, which interact with each other and share weights. DualGenerator introduces a local refinement module in generation paths, which captures general structures from partial inputs, and then refines shape details of the point cloud. It promotes completion in the unknown region and makes a distinction between different parts more obvious. Moreover, we design DGStyleGAN to improve the generation quality further. It promotes the robustness of this network combined with fusion analysis of dual-path completion results. Qualitative and quantitative evaluations demonstrate that our method is superior on MVP and Completion3D datasets. The performance will not degrade significantly after adding noise interference or sparse sampling.Keywords
All Related Versions
This publication has 9 references indexed in Scilit:
- Point Clouds Outlier Removal Method Based on Improved Mahalanobis and CompletionIEEE Robotics and Automation Letters, 2022
- Partial-to-Partial Point Generation Network for Point Cloud CompletionIEEE Robotics and Automation Letters, 2022
- PMP-Net++: Point Cloud Completion by Transformer-Enhanced Multi-Step Point Moving PathsIEEE Transactions on Pattern Analysis and Machine Intelligence, 2022
- FinerPCN: High fidelity point cloud completion network using pointwise convolutionNeurocomputing, 2021
- PointLIE: Locally Invertible Embedding for Point Cloud Sampling and Recovery2021
- Point Set Voting for Partial Point Cloud AnalysisIEEE Robotics and Automation Letters, 2021
- Multi-stage point completion network with critical set supervisionComputer Aided Geometric Design, 2020
- ECG: Edge-aware Point Cloud Completion with Graph ConvolutionIEEE Robotics and Automation Letters, 2020
- Learning Localized Representations of Point Clouds With Graph-Convolutional Generative Adversarial NetworksIEEE Transactions on Multimedia, 2020