STRIP: A Defence Against Trojan Attacks on Deep Neural Networks.
Yansong GaoChang XuDerui WangShiping ChenDamith Chinthana RanasingheSurya NepalPublished in: CoRR (2019)
Keyphrases
- neural network
- national security
- countermeasures
- fuzzy logic
- fuzzy systems
- chosen plaintext
- terrorist attacks
- pattern recognition
- artificial neural networks
- back propagation
- neural network model
- multilayer perceptron
- multi layer
- watermarking scheme
- feed forward
- deep learning
- neural nets
- self organizing maps
- fault diagnosis
- associative memory
- malicious attacks
- watermarking method
- security threats
- web services
- genetic algorithm