The project aim is twofold:
i) develop robust defenses against deepfakes both for identity document images and selfie-videos, and
ii) take advantage of the capabilities of generative AI to improve the state of the art in digital authentication by making algorithms more robust and balanced.