Scientists at La Sapienza University of Rome have created a new technology that can track and identify people through walls using just Wi-Fi signals. The system is called “WhoFi,” and it uses regular Wi-Fi routers to detect how a person’s body changes the signal. By studying those changes, the system can recognize each individual with up to 95.5 percent accuracy.
The WhoFi system does not need cameras, phones, or wearable devices. It only relies on how the body affects the path of a Wi-Fi signal. The team behind the research described the system as a new kind of biometric identifier that doesn’t use visuals or touch, but instead detects small differences in the way Wi-Fi waves bounce off each person’s body.
They wrote, “The core insight is that as a Wi-Fi signal propagates through an environment, its waveform is altered by the presence and physical characteristics of objects and people along its path. These alterations contain rich biometric information.”
To test their idea, scientists used a deep learning AI system called a transformer model. This model is able to study small changes in the Wi-Fi signals and recognize which person is causing them.
They tested it using a public dataset called NTU-Fi and tried it on 14 different people in different clothes and different places. The system could recognize people with an accuracy of up to 95.5 percent, even through walls and in places with poor lighting.
Researchers developed a system called WhoFi as an improvement over EyeFi. Previously, EyeFi launched in 2020 and reached only 75 percent accuracy. Now, with WhoFi, researchers are using Wi-Fi for human sensing.
At the same time, the idea of using Wi-Fi for detection is not new. In fact, scientists have studied it for years. Over time, they have used it to detect movement, gestures, and even falls. However, identifying people through walls without cameras or devices introduces a new capability. Consequently, this development also raises concerns among experts.
On the other hand, experts warn about serious privacy risks. For instance, people can see cameras or RFID tags and choose to avoid them. In contrast, Wi-Fi works differently because it surrounds us all the time. As a result, people rarely treat it as a tracking tool. Moreover, the system can operate silently in the background. Therefore, this makes detection difficult for users to notice.
A privacy expert said, “This research opens up new possibilities for non-invasive monitoring, but also for covert surveillance.” The expert added that irresponsible use could enable passive re-identification. A person may not know or consent to such tracking.
Researchers acknowledge these concerns. They said WhoFi does not directly collect personal identity data. It also does not reveal a person’s identity on its own. However, they agreed that strong rules must control such accurate tracking systems. These rules must prevent misuse in real-world applications.
The WhoFi project remains a research study for now. No company plans to use it in commercial products yet. Governments also have no plans for surveillance deployment. However, scientists warn about future risks. Smart Wi-Fi networks in buildings and cities may enable similar tracking in daily life.
Researchers said, “We believe WhoFi can be used in a responsible way.” They also said privacy protection remains necessary. They stressed that society must set limits before wide adoption begins.
The full research paper is currently available as a preprint on arXiv.
