Face Recognition (FR) in public space grows, way too much. The main examples are the pervert meta glasses, or simply securitarian-vibe massive deployment (for example the Flock camera, or Palantir collaborations across agencies and in the world). Biometric surveillance is constantly increasing and there is not a capable movement effective to defeat them. There are plenty of partially successful actions many actors, art pieces, policy efforts, technical research… but the surveillance state is winning, in many ways, in the political acceptance and the technology penetration. There are plenty of old and new anti-FR techniques: pain your face, shoot laser at cameras, wear fabric with special patterns, 3D-printed objects for your head, hat with faces on top and on the back… but what’s really working? we can’t know! Except by establishing a systematic, worldwide test and collection of successful mechanisms! Defeat Face Recognition system isn’t something we can do with a standalone technique. The “attack” is asymmetric, it evolves, it might utilize unknown models, and you never know what’s running behind a camera. Ghostmaxxing wants to play there, with a feedback loop, it’s free software and oped data, designed this way: Experiment locally. Use the browser lab to test makeup and other visual interventions against face recognition. Experiments remain in the browser unless you choose to upload them. Has been tried to find recurring “patterns”, but they mostly work when you talk to a Visual Model and you ask to mod a picture for you. Document what works! (With the participant’s consent), a success can be saved/shared/and re-produced a Ghostyles. (which is a javascript capable of casting an overlay in realtime over a face - check the example plugin). Successful camouflage (via images, before-after, or videos) can be submitted to the community gallery. Every submission is reviewed by a human moderator