You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
A defense algorithm which utilizes the combination of an auto- encoder and block-switching architecture. Auto-coder is intended to remove any perturbations found in input images whereas block switching method is used to make it more robust against White-box attack. Attack is planned using FGSM model, and the subsequent counter-attack by the prop…
This github repository contains the official code for the papers, "Robustness Assessment for Adversarial Machine Learning: Problems, Solutions and a Survey of Current Neural Networks and Defenses" and "One Pixel Attack for Fooling Deep Neural Networks"