Researchers create WiFi device able to detect people through walls
Researchers at Carnegie Mellon University, Pennsylvania, have created a method for accurately detecting the shape and movement of human bodies through walls using WiFi routers. Is this a privacy nightmare waiting to happen?
Remember Lucius Fox’s super invasive sonar technology in the finale of The Dark Knight? Well this is eerily similar, except that it’s real.
Scientists have reportedly developed technology able to detect the precise shape and movement of human bodies through walls using simple WiFi routers. Eerie, right?
Building on technology originally made by Facebook AI researchers called DensePose — which maps pixels of bodies in photos to create accurate 3D representations — scientists at Carnegie Mellon University, Pennsylvania, decided to pursue an entirely different aim based on security.
Outlined in the public journal arXiv, the researchers explained how taking the principle of DensePose and applying it to their deep neural networks allowed them to detect living 3D silhouettes of people through walls.
The mere presence of a WiFi connection is sufficient to kick this neural network into gear, shortly establishing where everyone in a room is and what they’re doing in real time.
Phase and amplitude signals sent to a regular internet router fire around the room and map out coordinates of specific body parts as they collide with them. Viewed on a monitor, the results are surprisingly detailed.
While this is undoubtedly an upping of the ante in terms of invasiveness, monitoring people without using cameras or expensive LiDAR software has, for some reason, been an elusive goal for tech experts since 2013.
The researchers wrote that they believe WiFi signals can ‘serve as a ubiquitous substitute’ for commercial CCTV cameras such as Nest or Ring, which may struggle with common obstacles like inadequate lighting and distortion.
Undeterred by the ongoing ethics debate surrounding intrusive technology and data confidentiality — which DensePose’s parent, Facebook, already has sketchy track record with — the university team actually believes its product is a positive advancement for privacy rights. Interesting.
‘Most households in developed countries already have WiFi at home, and this technology may be scaled to monitor the well-being of elderly people or just identify suspicious behaviours at home,’ a spokesperson declared.
The more obvious stance one might take, is that such technology going commercial would be way too risky for the masses.
The researchers are all too keen to point out that their neural network only requires our everyday WiFi routers to function, but that fact in itself is where the glaring oversight presents itself.
We’re regularly reminded by internet service providers to improve and update our security, and encouraging new pathways to our devices is completely counter-intuitive to that.
We’ve borne witness too many times to the ways in which cyber criminals can tamper with our data and devices, and this technological feat seems to bring far more cons than pros.
As we previously stated, however, this development will represent a significant victory for portions of the tech community. Whether that’s a cause for concern is entirely up to you.
Originally written by Jamie Watts for Thred.