This episode is an introduction to adversarial examples, a sort of "optical illusions" able to fool machine learning models. We will discuss what are adversarial examples, how can we generate them, why do they exist and we will explore if we can avoid them.  

Ver todos los episodios