Adversarial Tracker (Work-in-progress)

This repository strives to be a comprehensive zoo of all publicly-known adversarial attacks, and the state of their corresponding defences (if any). The objective is to allow ML engineers to be able to check this list and mitigate risk of adversarial attacks before deploying machine learning models to the wild.