RPG: Learning Recursive Point Cloud Generation

In this paper we propose a novel point cloud generator that is able to reconstruct and generate 3D point clouds composed of semantic parts. Given a latent representation of the target 3D model, the generation starts from a single point and gets expanded recursively to produce the high-resolution point cloud via a sequence of point expansion stages. During the recursive procedure of generation, we not only obtain the coarse-to-fine point clouds for the target 3D model from every expansion stage, but also unsupervisedly discover the semantic segmentation of the target model according to the hierarchical/parent-child relation between the points across expansion stages. Moreover, the expansion modules and other elements used in our recursive generator are mostly sharing weights thus making the overall framework light and efficient. Extensive experiments are conducted to show that our point cloud generator has comparable or even superior performance on both generation and reconstruction tasks in comparison to various baselines, and provides the consistent co-segmentation among instances of the same object class.

Authors

• Wei-Jan Ko
• Chen-Yi Chiu
• Yu-Liang Kuo
• Wei-Chen Chiu

Keywords

• Point cloud compression
• Solid modeling
• Three-dimensional displays
• Annotations
• Semantic segmentation
• Semantics
• Generators
• Point Cloud
• Point Cloud Generation
• 3D Point
• Latent Representation
• 3D Point Cloud
• Sequential Stages
• Expansion Stage
• Reconstruction Task
• Performance Comparison
• Scaling Factor
• Recurrent Neural Network
• Model Size
• Representative Structures
• Supplementary Video
• 3D Data
• 3D Shape
• Training Efficiency
• Variational Autoencoder
• Part Segmentation
• Point Cloud Data
• Hierarchical Representation
• Qualitative Examples
• Graph Convolution
• Point Cloud Reconstruction
• Smaller Model Size
• Object Instances
• Chamfer Distance
• Point Cloud Segmentation
• Airplane