Improving Semantic Embedding Consistency by Metric Learning for Zero-Shot Classiffication.
Maxime BucherStéphane HerbinFrédéric JuriePublished in: ECCV (5) (2016)
Keyphrases
- metric learning
- maximum variance unfolding
- distance metric
- distance metric learning
- machine learning and pattern recognition
- dimensionality reduction
- pairwise
- semi supervised
- nonlinear dimensionality reduction
- learning tasks
- multi task
- semi supervised clustering
- semi supervised learning
- feature space
- distance function
- semantic information
- mahalanobis metric
- machine learning
- subject to linear constraints
- labeled data
- image classification
- domain knowledge
- object recognition
- neural network
- data sets