no code implementations • 20 Jun 2020 • Esha Sarkar, Hadjer Benkraouda, Michail Maniatakos
In this work, we demonstrate that specific changes to facial characteristics may also be used to trigger malicious behavior in an ML model.
