Kernel Selection using Multiple Kernel Learning and Domain Adaptation in Reproducing Kernel Hilbert Space, for Face Recognition under Surveillance Scenario.
Samik BanerjeeSukhendu DasPublished in: CoRR (2016)
Keyphrases
- reproducing kernel hilbert space
- domain adaptation
- multiple kernel learning
- kernel methods
- semi supervised learning
- face recognition
- kernel function
- labeled data
- learning problems
- semi supervised
- cross domain
- linear combination
- multiple sources
- unlabeled data
- sentiment classification
- target domain
- feature selection
- machine learning
- feature space
- multi task
- face images
- feature representation
- supervised learning
- kernel matrix
- training data
- active learning
- transfer learning
- unsupervised learning
- document classification
- co training
- learning tasks
- gaussian process
- distance measure
- feature extraction
- test data