Improving Semantic Embedding Consistency by Metric Learning for Zero-Shot Classification.
Maxime BucherStéphane HerbinFrédéric JuriePublished in: CoRR (2016)
Keyphrases
- metric learning
- distance metric learning
- nearest neighbor classification
- distance metric
- margin maximization
- pattern recognition
- feature space
- maximum variance unfolding
- machine learning and pattern recognition
- semi supervised
- pairwise
- image classification
- fully supervised
- classification accuracy
- machine learning
- nonlinear dimensionality reduction
- dimensionality reduction
- supervised learning
- support vector
- feature set
- support vector machine
- kernel learning
- decision trees
- feature extraction
- face recognition
- learning tasks
- class labels
- distance function
- machine learning algorithms
- distance measure
- active learning
- feature vectors