---
title: "Invisible Camera: Karlsruhe Researchers Learn to Recognize People via Ordinary Wi-Fi with Nearly 100% Accuracy"
description: "Karlsruhe researchers have developed a system that recognizes people from ordinary Wi-Fi signals with nearly 100% accuracy. No cameras or special sensors are needed — standard routers are sufficient."
date: 2026-08-19T23:44:50.000Z
lang: en
url: https://xab.info/en/posts/invisible-camera-karlsruhe-wifi-people-recognition-nearly-100-accuracy
tags: [wifi, privacy, karlsruhe-institute-of-technology, ieee-802-11bf, surveillance, cybersecurity]
publisher: "XAB.info"
---

# Invisible Camera: Karlsruhe Researchers Learn to Recognize People via Ordinary Wi-Fi with Nearly 100% Accuracy

![Glowing orange Wi-Fi logo on a dark background — the technology enabling Karlsruhe scientists to recognize people via ordinary Wi-Fi with nearly 100% accuracy](https://xab.info/media/2026/08/20/wifi-identifikaciya-lyudey-pochti-100-tochnost-bez-kamer/wifi-identifikaciya-lyudey-pochti-100-tochnost-bez-kamer-1.webp)

## 🎯 Key Points

- The KASTEL Institute (KIT) proposed identifying people by analyzing Wi-Fi signals with nearly 100% accuracy on a sample of 197 volunteers.
- The technology uses unencrypted feedback signals from client devices to the router and requires no cameras or special equipment.
- Turning off a personal device does not protect against surveillance if other Wi-Fi devices are present in the area.
- To protect privacy, the IEEE 802.11bf standard is being developed.

The KASTEL Institute at the Karlsruhe Institute of Technology (KIT) has presented a system capable of identifying people inside a room without a single camera and without any special sensors. The approach is based on analyzing ordinary radio signals from wireless Wi-Fi networks. An experiment involving 197 volunteers demonstrated recognition accuracy of nearly 100 percent, with the result independent of both the angle at which the person was located and whether they were moving at the time of observation.

### How Wi-Fi Became an Invisible Camera

According to the lead researcher, Professor Thorsten Strufe, by observing the propagation of radio waves, the system is able to construct an image of the surrounding environment and the people present in it. “It works similarly to an ordinary camera; the only difference is that in our case, identification uses not light but ordinary radio waves,” the scientist explained. Thus, a familiar wireless network becomes a tool for visualizing space, requiring no optical sensors.

### Working Principle: Feedback Signals

The technology relies on feedback signals that client devices connected to the network send to the router. The key point: this information is transmitted without encryption, meaning that any participant within the network's coverage area can theoretically intercept and read it. By analyzing these radio signals over time, the researchers learned to reconstruct images of people from different viewpoints inside a room. Moreover, no special equipment is required — standard, unmodified Wi-Fi devices are sufficient.

### Experiment Results: 197 Volunteers

During testing, the system was evaluated on 197 volunteers and showed nearly complete identification accuracy. The method proved robust to changes in scene geometry: the viewing angle and the person's motion state (static or moving) did not reduce the result. This makes the technology potentially applicable in a wide range of real-world conditions — from a residential room to an open office space.

### Why Turning Off a Device Won't Help

An important property of the approach: a person does not need to carry a smartphone, smartwatch, or any other device with a Wi-Fi module. If such a device is present, turning it off does not protect against surveillance — the system works on the signals emitted by the other connected devices within the coverage area. This fundamentally distinguishes the threat from familiar scenarios where simply switching off a personal gadget is enough.

### Privacy Threat and Invisible Surveillance

Wireless networks are ubiquitous: they are present in homes, offices, restaurants, cafés, and other public places, creating a systemic threat to privacy. Traditionally, intelligence agencies and cybercriminals conduct surveillance by tapping into video surveillance systems and video doorbell cameras. However, wireless networks have an important advantage for the observer — they are invisible and raise no suspicion, since they are perceived by those around as ordinary infrastructure.

### Protection: The IEEE 802.11bf Standard

To protect people from such invisible surveillance, the researchers propose embedding privacy safeguards directly into the technology at the standard level. In particular, a version of the IEEE 802.11bf standard is currently being developed to address privacy concerns in the operation of wireless networks. The formation of such normative mechanisms could become a key barrier against the mass abuse of the ability to visualize people from radio signals.

## 🔍 Fact-Check Verification

- [People Have Been Learned to Recognize via Ordinary Wi-Fi — Nearly 100% Accuracy Without Cameras or Special Sensors window-new](https://3dnews.ru/1147098/lyudey-nauchilis-raspoznavat-po-obichnomu-wifi-pochti-100-tochnost-bez-kamer-i-spetsialnih-datchikov) - Подтверждены: институт KASTEL при KIT, 197 добровольцев, почти 100% точность, нешифрованные сигналы обратной связи, цитата проф. Торстена Штруфе, разработка IEEE 802.11bf.

## ❓ FAQ

### Q: Are cameras or special sensors required for the system to work?
**A:** No. The system operates on standard, unmodified Wi-Fi devices by analyzing radio signals, and requires no cameras or special equipment.

### Q: How many volunteers took part in the experiment, and what was the accuracy?
**A:** 197 volunteers participated in the experiment, and identification accuracy was nearly 100%, regardless of the person's angle or movement.

### Q: Will turning off my smartphone or watch help?
**A:** No. A person does not need to have a Wi-Fi device, and if they do, turning it off does not help — signals from the other devices in the network area are sufficient.

### Q: How is privacy proposed to be protected?
**A:** The researchers propose embedding privacy safeguards into the technology itself at the standard level, in particular through the IEEE 802.11bf standard currently under development.