QuEST: Quantized Embedding Space for Transferring Knowledge.
Himalaya JainSpyros GidarisNikos KomodakisPatrick PérezMatthieu CordPublished in: ECCV (21) (2020)
Keyphrases
- embedding space
- transferring knowledge
- transfer learning
- euclidean space
- low dimensional
- manifold learning
- knowledge transfer
- dimensionality reduction
- graph embedding
- input space
- high dimensional
- data points
- geometric structure
- multiple tasks
- geodesic distance
- labeled data
- learning tasks
- knowledge sharing
- high dimensional data
- principal component analysis
- active learning
- euclidean distance
- reinforcement learning
- learning algorithm
- vector space
- data sets
- unlabeled data
- collaborative filtering
- learning process
- image retrieval
- cross domain
- feature space
- training data
- feature extraction