Enabling Data Diversity: Efficient Automatic Augmentation via Regularized Adversarial Training.
Yunhe GaoZhiqiang TangMu ZhouDimitris N. MetaxasPublished in: CoRR (2021)
Keyphrases
- objective function
- data sets
- raw data
- synthetic data
- data sources
- image data
- statistical analysis
- database
- data processing
- complex data
- original data
- data collection
- knowledge discovery
- training data
- training samples
- sensor data
- data distribution
- small number
- prior knowledge
- genetic algorithm
- attribute values
- test data
- data mining
- neural network
- databases
- labelled data