Invisible Backdoor Attacks on Deep Neural Networks Via Steganography and Regularization.
Shaofeng LiMinhui XueBenjamin Zi Hao ZhaoHaojin ZhuXinpeng ZhangPublished in: IEEE Trans. Dependable Secur. Comput. (2021)
Keyphrases
- neural network
- pattern recognition
- countermeasures
- watermarking scheme
- information hiding
- recurrent neural networks
- genetic algorithm
- information security
- neural network model
- regularization parameter
- spread spectrum
- deep learning
- data hiding
- multi layer
- back propagation
- fuzzy logic
- image watermarking
- security protocols
- copyright protection
- digital images
- artificial neural networks
- hidden information
- high capacity
- secret information