An artificial intelligence (AI) technology capable of detecting illegally-installed hidden cameras (spy cams) using smartphones and inexpensive light-emitting diodes (LEDs) has been developed. With its low cost and high accuracy, this technology is expected to be effective in preventing illegal filming crimes in everyday spaces such as hotels and shared accommodations.


Detecting Spy Cams with Smartphones: "94% Accuracy at Low Cost" View original image

KAIST announced on August 30 that a research team led by Professor Han Jun from the School of Computing, in collaboration with the National University of Singapore (NUS) and Singapore Management University (SMU), has developed a detection technology called "SweepLED" with these capabilities.


SweepLED can be utilized by analyzing patterns of reflected light detected on the surface of objects as an LED light source, attached to a stationary smartphone camera, changes direction.


In the case of ordinary glossy objects, the location of the reflection spot changes or disappears depending on the direction of the light. However, due to the structure of the internal lens, aperture, and image sensor, the reflected shape on a camera lens deforms in a uniquely distinct way. SweepLED works by learning these patterns of reflection changes through video analysis based on deep learning, enabling it to distinguish camera lenses.


Unlike conventional detection methods that rely on the user's vision to find hidden cameras by simply spotting "bright dots," this approach utilizes AI to comprehensively analyze movements and shape changes in reflections that appear under various lighting angles. This allows it to pinpoint the location of hidden cameras.


AI-generated image. KAIST

AI-generated image. KAIST

View original image

The joint research team explained that using SweepLED based on this principle makes it more stable and convenient to locate camera lenses hidden inside objects commonly found in accommodations, such as chargers, clocks, remote controls, and decorations.


In an actual performance evaluation conducted on 30 different objects, the time required to inspect a single item was less than five seconds, and the detection accuracy recorded 94%.


Above all, the research team noted that the cost of key components for the LED case, which attaches to smartphones, is less than 10,000 won. This is expected to be advantageous for the future development and commercialization of affordable, smartphone accessory-type detectors and consumer products.


Professor Han Jun said, "Illegal filming using hidden cameras is a serious crime that threatens personal safety and privacy in everyday spaces. This study is significant in that it combines low-cost smartphone-based hardware with AI analysis, enabling even non-experts to easily utilize spy cam detection technology."



Meanwhile, Junghyuk Yoon, a doctoral student at KAIST, participated as the first author in this study. The research results were presented at the international academic conference on mobile computing, 'ACM MobiSys 2026,' held recently.


This content was produced with the assistance of AI translation services.

© The Asia Business Daily. All rights reserved. Unauthorized AI training and use prohibited.

Today’s Briefing