Effect of large-scale pre-training on full and few-shot transfer learning for natural and medical images.
Mehdi ChertiJenia JitsevPublished in: CoRR (2021)
Keyphrases
- medical images
- transfer learning
- learning tasks
- anatomical structures
- medical imaging
- training and test data
- knowledge transfer
- transfer knowledge
- labeled data
- deformable models
- medical image analysis
- real world
- mr images
- active learning
- machine learning
- supervised learning
- reinforcement learning
- semi supervised learning
- text categorization
- cross domain
- multi task
- transferring knowledge
- training examples
- text classification
- training set
- machine learning algorithms
- target domain
- video sequences
- domain adaptation
- image segmentation
- decision trees
- segmentation of medical images
- visual features
- text mining
- collaborative filtering
- test set
- video data
- unlabeled data
- image classification
- semi supervised