
As time progresses, my intense wish to become unnoticeable keeps increasing. However, the pursuit of complete invisibility, both online and offline, conflicts with my goals—especially since I cannot continue writing without an audience. It seems becoming a cryptid in the wilderness is off the table now, thanks to recent discoveries by researchers in Italy.
A team from La Sapienza University of Rome has created ‘Who-Fi’, a system that tracks individuals with surprising accuracy by monitoring how their bodies interact with Wi-Fi signals. By analyzing how one’s physical presence disrupts this vast network, researchers claim they can effectively ‘fingerprint’ individuals, achieving up to 95.5% accuracy. Does anyone else feel trapped in their physical form today, or is it just me?
The related research paper elaborates on how various biomarkers are extracted from Channel State Information (CSI) and processed via a Deep Neural Network (DNN) with a Transformer-based encoder.
This ‘Who-Fi’ method could serve as an alternative to traditional visual tracking methods, which can be obstructed by lighting, angles, or occlusion. Although now it feels somewhat suffocating to think that Wi-Fi signals can penetrate walls. This new approach has already found benign applications in various settings, such as detecting patient respiration and adjusting PC power states.
One can only imagine the potential hauntings of this tech if it were to be misused. Fortunately, this technology has not yet been adopted in practical scenarios. Although the research team efficiently implemented their project with just two TP-Link N750 routers, it is too early to dive into paranoia.
The remarkable accuracy purportedly links up with a rather modest sample size of ‘14 different subjects,’ each yielding ‘60 samples’ during short walks within defined testing areas. Consequently, ‘Who-Fi’ is more intriguing than ominous for now—at least until its capabilities evolve further.