/AdversarialAttacks

Designing a Black Box Adversarial Attack to deceive the efficiently trained model using Zeroth order SGD, SignSGD and Adam and comparison of results between them

Primary LanguageJupyter NotebookMIT LicenseMIT

Watchers