FGCN: Deep feature-based graph convolutional network for semantic segmentation of urban 3D point clouds

Keyphrases

• 3D Point Cloud 100%
• Semantic Segmentation 100%
• Feature-based 100%
• Graph Convolutional Network 100%
• Preprocessing Techniques 25%
• Benchmark Dataset 25%
• Challenging Tasks 25%
• Outdoor Scenes 25%
• Three-dimensional Space 25%
• 3D Point 25%
• Segmentation Map 25%
• Convolutional Neural Network 25%
• Feature Extracting 25%
• Indoor Scene 25%
• Voxelization 25%
• Neighborhood Relation 25%
• Semantic Scene Parsing 25%
• Spatial Arrangement 25%
• Part Segmentation 25%
• Urban Classification 25%
• Convolutional Neural Network Architecture 25%
• Object Structure 25%
• End-to-end Architecture 25%
• 2D Convolutional Neural Network 25%
• Memory Constraints 25%
• Segmentation Classification 25%
• Proposed Methodology 25%
• Computational Constraints 25%
• Quantization Artifacts 25%

Computer Science

• Point Cloud 100%
• Image Segmentation 100%
• Graph Convolutional Network 100%
• Deep Feature 100%
• Convolutional Neural Network 75%
• Preprocessing Step 25%
• Neural Network Architecture 25%
• Parsing 25%
• Extracted Feature 25%
• Segmentation Map 25%
• Neighborhood Relation 25%
• Graph Convolution 25%