CEREALS - Cost-Effective REgion-based Active Learning for Semantic Segmentation.
Radek MackowiakPhilip LenzOmair GhoriFerran DiegoOliver LangeCarsten RotherPublished in: CoRR (2018)
Keyphrases
- cost effective
- semantic segmentation
- active learning
- street scenes
- conditional random fields
- superpixels
- label transfer
- low cost
- object categories
- scene classification
- training examples
- weakly supervised
- cost effectiveness
- object class
- image segmentation
- semi supervised
- unlabeled data
- machine learning
- supervised learning
- object classes
- pascal voc
- learning algorithm
- object recognition
- object detection
- training set
- image retrieval
- labeled data
- long range
- learning process
- relevance feedback
- image understanding
- multiscale
- higher order
- data sets