Knowledge Enhanced Machine Learning Pipeline
                against Diverse Adversarial Attacks

            Nezihe Merve Gurel 1 * Xiangyu Qi 2 * Luka Rimanic 1 Ce Zhang 1 Bo Li 3

              Abstract              1. Introduction
                                Recent studies show that machine learning (ML) models are
                                vulnerable to different types of adversarial examples, which
    Despite the great successes achieved by deep neu-    are adversarially manipu ...