Researchers at La Sapienza University of Rome have developed a system that can identify individuals using only Wi-Fi signals, eliminating the need for cameras, smartphones, or wearables.
The system, called WhoFi, was introduced by computer scientists Danilo Avola, Daniele Pannone, Dario Montagnini, and Emad Emam in a study titled WhoFi: Deep Person Re-Identification via Wi-Fi Channel Signal Encoding.
According to the paper, WhoFi works by analyzing how the human body distorts electromagnetic waves, generating a unique biometric signature. Unlike traditional identification methods, it does not rely on visual input or physical contact.
Specifically, the system captures variations in Channel State Information (CSI)—such as changes in signal amplitude and phase—as Wi-Fi signals pass through or bounce off a person.
These subtle distortions are analyzed by a neural network trained to re-identify individuals with up to 95.5 percent accuracy using the NTU-Fi dataset, a benchmark developed by Nanyang Technological University in Singapore.
“The waveform is altered by the presence and physical characteristics of objects and people,” the researchers wrote. These distortions, they explained, carry distinct biometric information that allows the system to recognize a person over time.
The method builds on earlier efforts, such as the 2020 EyeFi system, which reached 75 percent accuracy.
For now, the technology remains confined to academic research, with no commercial or government use planned.
