[wlug] Fawkes: Protecting Privacy against Unauthorized Deep Learning Models

This paper <http://people.cs.uchicago.edu/%7Eravenben/publications/abstracts/fawkes-usenix20.html>, presented at USENIX Security 2020, discusses techniques for allowing you to continue to share images of yourself online while making them resistant to automated exploitation by “deep-learning” facial-recognition software. The researchers found that they could make subtle distortions to facial images, imperceptible to the human eye, that would severely confuse recognition services from major providers like Microsoft and Amazon. (Found from Bruce Schneier <https://www.schneier.com/blog/archives/2020/07/fawkes_digital_.html>.)