Siamese networks for generating adversarial examples

Mandar Kulkarni, Aria Abubakar

Machine learning models are vulnerable to adversarial examples. An adversary modifies the input data such that humans still assign the same label, however, machine learning models…