Hidden camera detectors miss nearly 60% of cameras –but a $7 AI phone case could solve that, and here's how...

A study from the University College of London revealed that a staggering 59% of the time, traditional hidden camera detectors fail to spot surveillance devices. This is because these detectors are often challenging to operate and interpret, with only 33% of hidden cameras being found when participants searched without a detector, improving slightly to 58.8% when a detector was used.

The standard method employed by these detectors is to shine light around the room, and it is up to the user to identify reflections from lenses. However, this can result in an emotional cost for users, as false positives may cause unnecessary anxiety. In response to these issues, a new device called SweepLED is being proposed as a solution.

Developed by KAIST and potentially constructible for $7, SweepLED is a small array of LEDs that move light and uses a phone's camera to capture a short video. This video is then analyzed by an AI model, achieving an impressive 94% detection accuracy based on the shape of the camera reflecting light differently. While the feasibility of turning this model into a viable product still needs to be determined, the collaboration of the National University of Singapore and Singapore Management University suggests that this idea is indeed promising.