RGCNN: Regularized Graph CNN for Point Cloud Segmentation

Abstract

Point cloud, an efficient 3D object representation, has become popular with the development of depth sensing and 3D laser scanning techniques. It has attracted attention in various applications such as 3D tele-presence, navigation for unmanned vehicles and heritage reconstruction. The understanding of point clouds, such as point cloud segmentation, is crucial in exploiting the informative value of point clouds for such applications. Due to the irregularity of the data format, previous deep learning works often convert point clouds to regular 3D voxel grids or collections of images before feeding them into neural networks, which leads to voluminous data and quantization artifacts. In this paper, we instead propose a regularized graph convolutional neural network (RGCNN) that directly consumes point clouds. Leveraging on spectral graph theory, we treat features of points in a point cloud as signals on graph, and define the convolution over graph by Chebyshev polynomial approximation. In particular, we update the graph Laplacian matrix that describes the connectivity of features in each layer according to the corresponding learned features, which adaptively captures the structure of dynamic graphs. Further, we deploy a graph-signal smoothness prior in the loss function, thus regularizing the learning process. Experimental results on the ShapeNet part dataset show that the proposed approach significantly reduces the computational complexity while achieving competitive performance with the state of the art. Also, experiments show RGCNN is much more robust to both noise and point cloud density in comparison with other methods. We further apply RGCNN to point cloud classification and achieve competitive results on ModelNet40 dataset.

References in corpus (9)

• Semi-Supervised Classification with Graph Convolutional Networks
• Spectral Networks and Locally Connected Networks on Graphs
• Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering
• Deep Convolutional Networks on Graph-Structured Data
• Learning Convolutional Neural Networks for Graphs
• VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection
• Dynamic Graph CNN for Learning on Point Clouds
• Accelerated filtering on graphs using Lanczos method
• SyncSpecCNN: Synchronized Spectral CNN for 3D Shape Segmentation

Cited by in corpus (27)

• Graph Convolutional Networks for Graphs Containing Missing Features
• PointContrast: Unsupervised Pre-training for 3D Point Cloud Understanding
• Relation-Shape Convolutional Neural Network for Point Cloud Analysis
• Self-Contrastive Learning with Hard Negative Sampling for Self-supervised Point Cloud Learning
• Two-Stream Graph Convolutional Network for Intra-oral Scanner Image Segmentation
• Deep Learning for 3D Point Clouds: A Survey
• DeltaConv: Anisotropic Operators for Geometric Deep Learning on Point Clouds
• 3D Point Cloud Descriptors in Hand-crafted and Deep Learning Age: State-of-the-Art
• PyramNet: Point Cloud Pyramid Attention Network and Graph Embedding Module for Classification and Segmentation
• PointGrow: Autoregressively Learned Point Cloud Generation with Self-Attention
• Geometric Back-projection Network for Point Cloud Classification
• DRINet++: Efficient Voxel-as-point Point Cloud Segmentation
• Point Attention Network for Semantic Segmentation of 3D Point Clouds
• Optimized Skeleton-based Action Recognition via Sparsified Graph Regression
• Task-Aware Sampling Layer for Point-Wise Analysis
• Graph Laplacian Regularized Graph Convolutional Networks for Semi-supervised Learning
• Revisiting 2D Convolutional Neural Networks for Graph-based Applications
• ePointDA: An End-to-End Simulation-to-Real Domain Adaptation Framework for LiDAR Point Cloud Segmentation
• Progressive Point Cloud Deconvolution Generation Network
• ASAP-Net: Attention and Structure Aware Point Cloud Sequence Segmentation
• Hierarchical Point Cloud Encoding and Decoding with Lightweight Self-Attention based Model
• Multi-task Graph Convolutional Neural Network for Calcification Morphology and Distribution Analysis in Mammograms
• Exploring Hypergraph Representation on Face Anti-spoofing Beyond 2D Attacks
• Learning to Segment 3D Point Clouds in 2D Image Space
• Airborne LiDAR Point Cloud Classification with Graph Attention Convolution Neural Network
• TSGCNet: Discriminative Geometric Feature Learning with Two-Stream GraphConvolutional Network for 3D Dental Model Segmentation
• PointNLM: Point Nonlocal-Means for vegetation segmentation based on middle echo point clouds