Heterogeneous domain adaptation using previously learned classifier for object detection problem.
Azadeh Sadat MozafariMansour JamzadPublished in: ICIP (2014)
Keyphrases
- domain adaptation
- previously learned
- transfer learning
- training data
- target domain
- labeled data
- training examples
- co training
- cross domain
- multiple sources
- semi supervised learning
- training samples
- test data
- knowledge transfer
- previously unseen
- semi supervised
- unlabeled data
- reinforcement learning
- learning tasks
- decision trees
- sentiment classification
- support vector
- domain theory
- multi task learning
- active learning
- class labels
- supervised learning
- reinforcement learning methods
- feature space
- document classification
- classification algorithm
- feature selection
- training set
- support vector machine
- domain specific
- principal component analysis
- data points