Self-adaptive logit balancing for deep neural network robustness: Defence and detection of adversarial attacks.
Jiefei WeiLuyan YaoQinggang MengPublished in: Neurocomputing (2023)
Keyphrases
- neural network
- attack detection
- false positives
- false alarms
- digital image watermarking
- detection method
- object detection
- artificial neural networks
- detection algorithm
- neural network model
- back propagation
- automatic detection
- countermeasures
- multi layer
- detection accuracy
- image watermarking scheme
- semi fragile watermarking
- neural network is trained
- network analysis
- neural nets
- self organizing maps
- regression model
- anomaly detection
- pattern recognition