Supervoxel-based and Cost-Effective Active Learning for Point Cloud Semantic Segmentation.
Shanding YeYongjian FuHu LinZhe YinZhijie PanPublished in: ITSC (2022)
Keyphrases
- cost effective
- point cloud
- semantic segmentation
- active learning
- conditional random fields
- superpixels
- structure from motion
- weakly supervised
- low cost
- scene classification
- point sets
- semi supervised
- object categories
- object classes
- object class
- supervised learning
- training examples
- learning algorithm
- pascal voc
- sparse reconstruction
- labeled data
- training set
- bounding box
- object recognition
- machine learning
- image set
- single image
- viewpoint
- feature extraction
- image segmentation
- feature selection
- data sets