ReDAL: Region-based and Diversity-aware Active Learning for Point Cloud Semantic Segmentation.
Tsung-Han WuYueh-Cheng LiuYu-Kai HuangHsin-Ying LeeHung-Ting SuPing-Chia HuangWinston H. HsuPublished in: ICCV (2021)
Keyphrases
- point cloud
- semantic segmentation
- active learning
- conditional random fields
- superpixels
- weakly supervised
- structure from motion
- scene classification
- object categories
- training examples
- point sets
- object classes
- semi supervised
- image segmentation
- learning algorithm
- machine learning
- training set
- object detection
- object class
- pascal voc
- multiscale
- sparse reconstruction
- object classification
- visual words
- image retrieval