How do AI systems decide on personal labels such as age and gender based on looks alone? Face / On is an interactive installation that explores what systems see. It takes the form of a closed door embedded with a facial recognition system. Face/On is a gatekeeper deciding on who looks like they fit its criteria, and who will be permitted entry past the threshold. It may be necessary to adopt a disguise to fool the system.
Video Credit: Dave Weir
Photo Credit: Ibi Feher, Control Shift
Approach the door
You will be allocated an age and gender
Only this persona is allowed through; a disguise may be necessary
The decisions are made by an AI. Where are the boundaries?
Use the available means to change your face
Here are a few ways to change your face using free browser based apps from the web. Some of them require a photo of your face. If you've already got one on your phone, you can use that or pop outside to take one without a mask on.
Face / On takes a picture every few seconds and stores it locally. Images are sent to Microsoft Azure Face API for analysis. No pictures will be used without permission.
Get in touch: email@example.com
Visual Design for Face / On has been develop with Copa Ipa. Metalwork by Tom Cherry. Produced with support from Control Shift. Thank you to Ruby Soho for lending your workshop and Professor Andrew Charlesworth for sharing your knowledge and ideas around facial recognition.