1 code implementation • 14 Nov 2021 • Roee Ben Shlomo, Yevgeniy Men, Ido Imanuel
Generating adversarial examples is the art of creating a noise that is added to an input signal of a classifying neural network, and thus changing the network's classification, while keeping the noise as tenuous as possible.
