Self-supervised Learning of Geometrically Stable Features Through Probabilistic Introspection.
David NovotnýSamuel AlbanieDiane LarlusAndrea VedaldiPublished in: CoRR (2018)
Keyphrases
- learning process
- prior knowledge
- learning algorithm
- neural network
- learning systems
- learning models
- supervised learning
- online learning
- feature representations
- low level
- semi supervised
- feature hierarchies
- learning scheme
- classification models
- uncertain data
- incremental learning
- learning tasks
- knowledge acquisition
- text classification
- classification accuracy
- bayesian networks
- similarity measure
- feature extraction