A bag-to-class divergence approach to multiple-instance learning.
Kajsa MøllersenJon Yngve HardebergFred GodtliebsenPublished in: CoRR (2018)
Keyphrases
- multiple instance learning
- single instance
- multiple instance
- positive bags
- diverse density
- image categorization
- multi class
- supervised learning
- class labels
- object based image retrieval
- image annotation
- semi supervised learning
- semi supervised
- concept learning
- visual features
- labeled data
- data sets
- density estimation
- object detection
- support vector
- multi instance
- reinforcement learning
- feature extraction
- computer vision