Generating fake fingerprints to unlock smartphones

Our Phd candidate Philip Bontrager won best paper at BTAS, the IEEE International Conference on Biometrics: Theory, Applications, and Systems.  Written together with GIL Director Julian Togelius, Aditi Roy, Nasir Memon and Arun Ross, the paper covers the vulnerability of fingerprint scanners to fake prints generated by an adversarial neural network.

Recently, the paper was tweeted out by the head of research at F-Secure, and after that it was picked up by the news media.
Wired: https://www.wired.com/…/deepmasterprints-fake-fingerprints…/
Gizmodo: http://www.gizmodo.co.uk/…/ai-can-now-fake-fingerprints-th…/
Guardian: https://www.theguardian.com/…/fake-fingerprints-can-imitate…
Motherboard: https://motherboard.vice.com/…/researchers-created-fake-mas…
Engadget: https://www.engadget.com/…/ai-fingerprints-biometric-scann…/

The paper itself can be found here: