Using adversarial images to improve outcomes of federated learning for non-IID data.
Anastasiya DanilenkaMaria GanzhaMarcin PaprzyckiJacek MandziukPublished in: CoRR (2022)
Keyphrases
- image data
- data sets
- raw data
- prior knowledge
- learning algorithm
- image classification
- high quality
- database
- input data
- data sources
- digital data
- image database
- small number
- data analysis
- learning process
- edge detection
- data mining techniques
- image analysis
- image retrieval
- missing data
- learning models
- user input
- multi agent
- image registration
- image features
- supervised learning
- spatial data
- image collections