AI Weekly Malaysia

Back to items Summaries

This ‘adversarial’ pattern can prevent surveillance cameras from detecting you

ID
12419
Status
summarized
Published
09 Aug 2026, 10:00 PM
Fetched
09 Aug 2026, 10:40 PM
Provider
TechCrunch
Category
technology
Original URL
https://techcrunch.com/2026/08/09/this-adversarial-pattern-can-prevent-surveillance-cameras-from-detecting-you/
Source URL
https://techcrunch.com/feed/

Summary

Score
4.0
Created
09 Aug 2026, 10:42 PM
Tags
Audience
ai_ml_learnersdevelopers

What happened

Bill Swearingen ran roughly 31 million tests to generate adversarial visual patterns that, when printed on clothing or vehicles, prevent common surveillance cameras and license plate readers from detecting the covered objects or people. He demonstrated the project, called noRecognition, at Def Con 2026 in Las Vegas, showing it can defeat real-world detection algorithms without blocking the cameras from recording video itself.

Why it matters

For AI/ML learners and builders working on computer vision, this is a concrete, real-world demonstration of adversarial examples against deployed surveillance systems—useful as a case study if you build or evaluate object-detection models, but not something most developers or founders need to act on immediately.

Discussion angle

How robust are the object-detection models you ship against physical adversarial patterns, and does this kind of evasion matter for any CV-based product or service you build?

Top