📰 Curated from TechCrunch
📖 Read full article→
This ‘adversarial’ pattern can prevent surveillance cameras from detecting you - TechCrunch
Technology and ScienceBy Zack Whittaker8/9/20261 min read
A security researcher has designed an algorithm that can create computer-generated patterns capable of hiding people, faces, and vehicles from detection by surveillance cameras.
✨ Key Highlights
- Cybersecurity professional Bill Swearingen developed "noRecognition," a project producing computer-generated "adversarial" patterns designed to prevent common surveillance cameras and license plate readers from detecting whatever the patterns cover, including people and vehicles.
- After roughly 31 million tests conducted over about a year, Swearingen says he can now generate these patterns on demand for application to clothing and objects.
- The patterns do not stop cameras from recording footage; instead, they scramble the detection algorithms so cameras fail to trigger automatic identification alerts, effectively making a person or object harder to single out.
- Swearingen publicly demonstrated the technique for the first time on Friday at the Def Con cybersecurity conference in Las Vegas, showing a pattern printed on a vehicle successfully defeating detection in a real-world setting.
- Framing privacy as a fundamental right, Swearingen said his goal is to let people "opt out of being tracked," noting concerns about pervasive surveillance and the government's use of driver's license photos for facial recognition.
- Based in Kansas City, where he co-founded the SecKC cybersecurity meetup, Swearingen said his worries deepened last year when he considered attending a protest and feared cameras could track people exercising their free expression rights.
Bill Swearingen has spent the past year running largely the same test, over and over again. The goal was to produce a computer-generated pattern that could block the surveillance cameras lining Ameri