FPS-Net: A convolutional fusion network for large-scale LiDAR point cloud segmentation

Keyphrases

• Multi-channel 100%
• Computational Speed 100%
• Projection Method 100%
• Semantic Segmentation 100%
• Convolutional 100%
• Fusion Network 100%
• Fused Features 100%
• Modality-specific Features 100%
• MIoU 100%
• RangeNet++ 100%
• LiDAR Point Cloud Segmentation 100%
• 2D Image 50%
• Performance Improvement 50%
• Encoder 50%
• Reliable Performance 50%
• Decoder 50%
• Projected Image 50%
• Output Space 50%
• Accuracy Improvement 50%
• High-dimensional Feature Space 50%
• Pixel-level Fusion 50%
• Single Input 50%
• Pixel Classification 50%
• Point Cloud Data 50%
• 3D Point Cloud 50%
• LiDAR Point Cloud 50%
• Scene Understanding 50%
• Point Cloud Segmentation 50%
• Autonomous Cars 50%
• Encoder-decoder 50%
• Convolution Block 50%
• Information Capacity 50%
• Residual Dense Block 50%
• Modality Fusion 50%
• SemanticKITTI 50%
• Recurrent Convolution 50%
• Channel Images 50%
• KITTI Benchmark 50%
• Spherical Projection 50%
• Multiple Receptive Fields 50%

Computer Science

• Point Cloud 100%
• Point Cloud Segmentation 100%
• Image Segmentation 66%
• Scene Understanding 33%
• Building-Blocks 33%
• Performance Improvement 33%
• Dimensional Feature Space 33%
• Multiple Channel 33%
• Receptive Field 33%
• Convolution 33%