This is nothing new. Public spaces have no right to privacy and you are allowed to record phone and video unless on private property or in special cases.
Literally grab any basic streaming camera, use opencv or yolo with ffmpeg to extract images from live video and find faces and crop faces into a picture to upload via yandex or pimeyes. Then return a confidence rating.
This has been around for years, total nothing burger of a headline except its a smaller form factor.
This is nothing new. Public spaces have no right to privacy and you are allowed to record phone and video unless on private property or in special cases.
Literally grab any basic streaming camera, use opencv or yolo with ffmpeg to extract images from live video and find faces and crop faces into a picture to upload via yandex or pimeyes. Then return a confidence rating.
This has been around for years, total nothing burger of a headline except its a smaller form factor.
Linked article is to the students' Google Doc explaining the architecture.
More here: https://www.404media.co/someone-put-facial-recognition-tech-...
Interesting project. Did they unintentionally create Glasshole 2.0 or is the bug the feature?
You’d think they’d add a second camera for depth perception and gestures.