Using Adversarial Machine Learning, Researchers Look to Foil Facial Recognition

Researchers at the National University of Singapore published a technique that locates the image areas where changes can best disrupt image-recognition algorithms, without being noticed by humans.