Jacobian norm with Selective Input Gradient Regularization for interpretable adversarial defense
Deyin Liu, Lin Wu, Bo Li, Farid Boussaïd, Mohammed Bennamoun, Xianghua Xie, Chengwu Liang
Topics & Concepts
InterpretabilityAdversarial systemComputer scienceJacobian matrix and determinantDeep neural networksArtificial intelligenceMachine learningRobustness (evolution)Regularization (linguistics)Deep learningNorm (philosophy)MathematicsPolitical scienceLawChemistryBiochemistryApplied mathematicsGeneAdversarial Robustness in Machine LearningExplainable Artificial Intelligence (XAI)Anomaly Detection Techniques and Applications