An experience / installation that exposes the algorithmic injustices and inconsistencies of face recognition software.

The foundation of my project is finding out how face recognition works. How it reads your face. I’ve been doing that by on one hand applying different filters and distortions on Mark Zuckerberg and on the other hand looking into applying those distortions on different people. This includes different race, age and gender. I am collecting all of this into a database to get a grip on how face recognition works but more specifically to find out when it doesn’t work.

To be able to hide from face recognition and also to expose how it works, I build a system that uses existing face recognition software to track the emotion, face data, age and gender of each person participating. After this data is collected a personal mask is created. This mask will be based on your face features but also on how you will interact with the screen. So things like emotion and time will influence your mask as well. Meaning that 1 person could have 6 different masks.

Once you press on submit all this personal data will be send into my realtime database. This database will collect everyones entries, showcasing the uncanniness of keeping such a database. But at the same time you will also be able to download your own personal mask from here.

After downloading your personal mask, you will be able to use it during your online video meetings and streaming sessions, and also on your preferred social media platform. For now the focus is on snapchat as it’s actually working on developing a virtual camera that can replace your default phone camera, allowing you to apply these filters in any app you would like.

enter mask database