KAIST Builds $7 SweepLED Add-on That Detects Hidden Cameras With 94% Accuracy
Updated
Updated · New Atlas · Sep 21
KAIST Builds $7 SweepLED Add-on That Detects Hidden Cameras With 94% Accuracy
3 articles · Updated · New Atlas · Sep 21
Summary
KAIST and Singapore partners built SweepLED, a smartphone-case prototype that detected hidden cameras with about 94% accuracy and inspected each object in roughly five seconds.
The $7 device tackles false alarms by sweeping 15 LEDs across a target and using AI to analyze how reflections change over time, rather than flagging any single bright spot.
Researchers tested the system on 30 objects, including camera-equipped items and ordinary household goods, after stabilizing video and comparing frames with LEDs on and off.
The project targets rising privacy concerns in hotels and short-term rentals, where hidden cameras remain a problem despite Airbnb's global ban on indoor security cameras.
SweepLED is still a prototype, but the team presented the work at ACM MobiSys 2026, pointing to a low-cost consumer privacy tool if commercialized.