Domain adaptation based on domain-invariant and class-distinguishable feature learning using multiple adversarial networks.
Cangning FanPeng LiuTing XiaoWei ZhaoXianglong TangPublished in: Neurocomputing (2020)
Keyphrases
- domain adaptation
- multiple sources
- covariate shift
- learning process
- semi supervised
- cross domain
- source domain
- reinforcement learning
- active learning
- unlabeled data
- co training
- knowledge transfer
- learning tasks
- test data
- labeled data
- prior knowledge
- learning algorithm
- natural language processing
- target domain
- feature space
- training data
- face recognition