A cybersecurity expert has developed an artificial intelligence system that generates special optical patterns designed to prevent surveillance cameras and recognition software from detecting people or vehicles.
The project, named noRecognition, uses reinforcement learning to train models that confuse computer vision algorithms. These patterns scramble identification capabilities without blocking the actual video recording process itself.
Following successful real-world demonstrations at a major cybersecurity conference, the creator aims to make these patterns available on everyday items like apparel and vehicle skins to help citizens protect their privacy.
- AI-generated patterns block automated surveillance detection
- System trained through millions of machine learning iterations
- Successfully tested in public against real tracking hardware
- Plans to release privacy-focused merchandise and vehicle skins
Sources:
