Attention-Based Multi-Scale Graph Convolution for Point Cloud Semantic Segmentation.
Perpetual Hope AkwensiRuisheng WangPublished in: IGARSS (2022)
Keyphrases
- point cloud
- semantic segmentation
- multiscale
- superpixels
- scene classification
- graph structure
- weakly supervised
- conditional random fields
- object categories
- structure from motion
- point sets
- image processing
- natural images
- scale space
- image representation
- object class
- sparse reconstruction
- visual attention
- object classes
- image set
- image classification
- information extraction
- pascal voc
- viewpoint
- video sequences
- image segmentation