New Adversarial Patterns Block Surveillance Cameras from Detecting Individuals

Bill Swearingen, a cybersecurity professional, has developed a project called noRecognition that creates computer-generated patterns to block surveillance cameras from detecting people or vehicles. These patterns scramble the ability of license plate readers and facial recognition systems to identify objects, making individuals ‘a needle in a haystack’ again. The technology was tested at Def Con 2026, where a vehicle covered with the pattern evaded detection by Flock cameras. Swearingen’s work builds on prior efforts to counter surveillance, using reinforcement learning models to refine patterns that defeat multiple camera algorithms. The project aims to provide high-quality, fashionable apparel and vehicle skins while keeping advanced patterns offline to prevent misuse. This innovation highlights growing concerns about privacy in an era of pervasive algorithmic surveillance.