PCN: Point Completion Network

Abstract

Shape completion, the problem of estimating the complete geometry of objects from partial observations, lies at the core of many vision and robotics applications. In this work, we propose Point Completion Network (PCN), a novel learning-based approach for shape completion. Unlike existing shape completion methods, PCN directly operates on raw point clouds without any structural assumption (e.g. symmetry) or annotation (e.g. semantic class) about the underlying shape. It features a decoder design that enables the generation of fine-grained completions while maintaining a small number of parameters. Our experiments show that PCN produces dense, complete point clouds with realistic structures in the missing regions on inputs with various levels of incompleteness and noise, including cars from LiDAR scans in the KITTI dataset.

arXiv comment3DV 2018 oral. Honorable mention for Best Paper award

References in corpus (9)

• Adam: A Method for Stochastic Optimization
• Joint 2D-3D-Semantic Data for Indoor Scene Understanding
• ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenes
• Dynamic Graph CNN for Learning on Point Clouds
• Dense 3D Object Reconstruction from a Single Depth View
• PointCNN: Convolution On $\mathcal{X}$-Transformed Points
• Shape Completion using 3D-Encoder-Predictor CNNs and Shape Synthesis
• Improved Adversarial Systems for 3D Object Generation and Reconstruction
• FoldingNet: Point Cloud Auto-encoder via Deep Grid Deformation

Cited by in corpus (7)

• PF-Net: Point Fractal Network for 3D Point Cloud Completion
• Patch-based Progressive 3D Point Set Upsampling
• PU-GCN: Point Cloud Upsampling using Graph Convolutional Networks
• PointGrow: Autoregressively Learned Point Cloud Generation with Self-Attention
• One Framework to Register Them All: PointNet Encoding for Point Cloud Alignment
• Recovering and Simulating Pedestrians in the Wild
• TreeGCN-ED: Encoding Point Cloud using a Tree-Structured Graph Network