Best Practices in Active Learning for Semantic Segmentation.
Sudhanshu MittalJoshua NiemeijerJörg P. SchäferThomas BroxPublished in: DAGM (2023)
Keyphrases
- semantic segmentation
- active learning
- street scenes
- conditional random fields
- scene classification
- object categories
- training examples
- label transfer
- superpixels
- pascal voc
- semi supervised
- weakly supervised
- object class
- learning algorithm
- object classes
- machine learning
- object detection
- object recognition
- unlabeled data
- supervised learning
- image understanding
- labeled data
- training set
- image retrieval
- decision trees
- object segmentation
- visual words
- learning process
- natural images
- small number