Towards robust neural networks via a global and monotonically decreasing robustness training strategy.
Zhen LiangTaoran WuWanwei LiuBai XueWenjing YangJi WangZhengbin PangPublished in: Frontiers Inf. Technol. Electron. Eng. (2023)
Keyphrases
- neural network
- training process
- training algorithm
- high robustness
- neural network training
- feedforward neural networks
- artificial neural networks
- computationally efficient
- back propagation
- multi layer perceptron
- computational efficiency
- pattern recognition
- backpropagation algorithm
- global information
- fuzzy systems
- data sets
- digital image watermarking
- neural network structure
- global image statistics
- error back propagation
- multilayer neural network
- selection strategy
- network architecture
- multi layer
- recurrent neural networks
- feed forward
- neural network model
- training examples
- sufficient conditions
- digital images
- support vector
- training data
- genetic algorithm