With deep learning applications blossoming, it is important to understand what makes these models tick. Here I demonstrate, using simple and reproducible examples, how and why deep neural networks can be easily fooled. I also discuss potential solutions.
Read more