Generating Adversarial Examples with Better Transferability via Masking Unimportant Parameters of Surrogate Model.
Dingcheng YangWenjian YuZihao XiaoJiaqi LuoPublished in: IJCNN (2023)
Keyphrases
- computational model
- parameter values
- parametric models
- neural network
- formal model
- measured data
- image segmentation
- generation process
- probabilistic model
- parameter estimation
- similarity measure
- mathematical model
- experimental data
- kalman filter
- sensitivity analysis
- analytical model
- control parameters
- maximum likelihood estimation
- linear model
- parameter identification
- bayesian framework
- process model
- theoretical framework
- theoretical analysis
- least squares
- probability distribution
- prior knowledge
- high level
- genetic algorithm