SIT: Stochastic Input Transformation to Defend Against Adversarial Attacks on Deep Neural Networks
Amira Guesmi, Ihsen Alouani, Mouna Baklouti, Tarek Frikha, Mohamed Salah Abid
Abstract
To better combat the impact of adversarial samples on deep neural networks, a model-agnostic stochastic input transformation (SIT) preprocessing technique is proposed in this article. The inputs are transformed into a new domain to minimize the impact of the adversarial perturbations.
Topics & Concepts
Adversarial systemPreprocessorTransformation (genetics)Computer scienceDeep neural networksDomain (mathematical analysis)Artificial intelligenceArtificial neural networkTheoretical computer scienceMathematicsChemistryBiochemistryMathematical analysisGeneAdversarial Robustness in Machine LearningAnomaly Detection Techniques and ApplicationsAdvanced Neural Network Applications