They’re known as deepfakes – photos or videos that have been very convincingly manipulated to depict people saying or doing things that they never actually said or did. They’re potentially quite the problem, so an experimental new deep neural network has been designed to spot them.
Comments are closed.