BEL: A Bag Embedding Loss for Transformer Enhances Multiple Instance Whole Slide Image Classification.
Daniel SensArio SadafiFrancesco Paolo CasaleNassir NavabCarsten MarrPublished in: ISBI (2023)
Keyphrases
- multiple instance
- image classification
- single instance
- instance level
- bag of words
- diverse density
- multiple instance learning
- visual features
- real valued
- concept learning
- learning problems
- semi supervised learning
- image features
- image representation
- regression problems
- supervised learning
- multi instance learning
- multi label
- feature extraction
- relational learning
- data sets
- maximum margin
- multi instance
- kernel methods
- feature space
- support vector
- reinforcement learning
- decision trees
- feature selection
- computer vision