Archive — US — techcrunch
This ‘adversarial’ pattern can prevent surveillance cameras from detecting you
Bill Swearingen developed patterns that hinder surveillance cameras from detecting objects and people. After 31 million tests, his 'noRecognition' project proved effective in real-world scenarios, allowing individuals to protect their privacy from algorithmic surveillance. The project debuted at Def Con 2023, demonstrating that these techniques could become a tool for opting out of constant tracking.
Key points
- Bill Swearingen conducted 31 million tests to create patterns that block surveillance detection.
- The patterns are designed to confuse detection algorithms used by cameras and license plate readers.
- Swearingen's project aims to protect privacy, allowing people to opt-out of surveillance.
- He demonstrated the effectiveness of his patterns at the Def Con cybersecurity conference.
- His approach builds upon previous methods to defeat detection systems and uses reinforcement learning for continuous improvement.
Why it matters
- As surveillance technology rapidly evolves, there is a growing need for privacy protections.
- The rise of facial recognition and license plate tracking has sparked debates about civil liberties.
- This project aligns with trends toward personal data protection, especially amid concerns over government surveillance.
Swearingen's innovations signal a significant step toward enhancing personal privacy in an era of increasing surveillance, reinforcing the importance of developing countermeasures against intrusive technologies.