How to augment a small learning set for improving the performances of a CNN-based steganalyzer?
Mehdi YedroudjMarc ChaumontFrédéric CombyPublished in: Media Watermarking, Security, and Forensics (2018)
Keyphrases
- small number
- learning process
- learning problems
- real time
- incremental learning
- graphical models
- cellular neural networks
- hypothesis space
- inductive inference
- learning scenarios
- learning tasks
- background knowledge
- supervised learning
- multi class
- probability distribution
- artificial neural networks
- reinforcement learning
- data sets