Paint Your Face Away (2020) is an art project that critically responds to the issues of privacy and surveillance concerning the biometric data economy of faces online. The project combines the methods of generative art, painting and adversarial attacks.
The project develops and makes accessible to the public a browser-based application of an online face painter. The painter creates portrait pictures that can resist the technology and operation of face scraping that powers some facial recognition systems.
Facial recognition is increasingly used for surveillance and can be trained with digitised facial images of individuals online. Such images can be collected (scraped) and used without the subjects’ consent. The trained models can be used to identify or track the individuals possibly without their knowledge.
On the painter, participants are asked to digitally paint over an image of their face until no face is found in it by a face detection algorithm. The public can utilise the tool to create portraits that can obfuscate machine vision and surveillance where the subject’s facial biometrics can be creatively erased or rendered useless as legitimate data by the digital paint. Moreover, the participants can contribute to the public database of adversarial self-portraits viewable on the website.
The web application was commissioned by and co-produced with Fotomuseum Winterthur.
Production support by Alejandro Daniel Ball from Agorama, Ashwin D’Cruz, Tobias Stenberg. Research advice by Agorama, Simon Crowe, Alexander Fefegha, Marco De MuTiis. The tools used for the current version of the application include: P5.js, ml5.js, face-api.js.
Access the painter from https://paintyourfaceaway.net
More information about the project: https://paintyourfaceaway.net/info