FACM: Correct the Output of Deep Neural Network with Middle Layers Features against Adversarial Samples.
Xiangyuan YangJie LinHanlin ZhangXinyu YangPeng ZhaoPublished in: CoRR (2022)
Keyphrases
- neural network
- feature values
- training examples
- unsupervised feature learning
- data sets
- neural network model
- feature set
- co occurrence
- feature extraction
- genetic algorithm
- classification accuracy
- image features
- low level
- high dimensional
- feature space
- false positives
- self organizing maps
- multi agent
- neural nets
- multi layer
- deep learning
- single layer
- desired output