Omni-Supervised Point Cloud Segmentation via Gradual Receptive Field Component Reasoning.
Jingyu GongJiachen XuXin TanHaichuan SongYanyun QuYuan XieLizhuang MaPublished in: CVPR (2021)
Keyphrases
- point cloud
- receptive fields
- biologically inspired
- natural images
- structure from motion
- image segmentation
- laser scanner
- visual cortex
- image representation
- point sets
- primary visual cortex
- multiscale
- feature selection
- supervised learning
- space time
- learning algorithm
- image regions
- higher order
- saliency map
- training samples
- low dimensional
- decision trees
- machine learning