Selecting Relevant Features from a Multi-domain Representation for Few-Shot Classification.
Nikita DvornikCordelia SchmidJulien MairalPublished in: ECCV (10) (2020)
Keyphrases
- multi domain
- selecting relevant features
- pattern recognition
- feature space
- support vector
- feature selection
- decision trees
- class labels
- feature extraction
- domain specific
- feature vectors
- image representation
- image classification
- classification accuracy
- training set
- data analysis
- data sets
- text classification
- generative model
- object recognition
- video data
- multiscale
- dimension reduction
- cross domain
- social networks
- learning algorithm