Science

New technology allows you to see through walls using Wi-Fi

⇧ [VIDÉO] You may also like this affiliate content (after ads)

Theoretically, our housing should be the most secluded place. But the more technology advances, the more that certainty becomes uncertainty. Not to mention that our movements, our personal data, almost everything can be found on the Internet. Recently, researchers at Carnegie Mellon University have developed a method for detecting the 3D shape and movements of human bodies in a room using only Wi-Fi routers.

Advances in computer vision and machine learning techniques have led to significant developments in 2D and 3D human position estimation using RGB (red, green, blue) cameras, LiDAR (laser assisted remote sensing) and radars. This is true, for example, in the military sphere or in the police.

However, the assessment of the position of human bodies in images is affected by occlusion, that is, a narrow or even opaque field of view, and poor lighting quality (darkness or glare), which are typical for many situations. Not to mention, radar and lidar technologies require specialized equipment that is expensive and energy intensive, and therefore difficult to use.

:: T-SHIRT THAT SUPPORTS SCIENCE! ::

Show the world your passion for space and that you also support the fight against global warming.

In addition, placing these sensors in closed places raises serious privacy concerns. To address these limitations, a recent study explored the use of Wi-Fi antennas (1D sensors) for body segmentation and detection of body key points.

In this context, a research team at Carnegie Mellon University addresses these issues by proposing a method for detecting the 3D shape and movements of human bodies indoors using only Wi-Fi routers in scenarios with a limited field of view and including multiple people. Their study is published on arXiv (and has not yet been peer-reviewed).

Entrance door: WiFi

In particular, the authors of the study used DensePose, a system that allows you to display all the pixels on the surface of the human body in a photograph. As detailed in the Vice article, DensePose was developed by London-based and Facebook AI researchers. They then developed a deep neural network that maps the phase and amplitude of Wi-Fi signals sent and received by routers to coordinates on the human body. The data is processed using computer vision algorithms.

You should be aware that the search for less expensive alternatives to lidar systems is outdated. In fact, in 2013, a group of MIT researchers found a way to use mobile phone signals to see through walls; in 2018, another MIT team used Wi-Fi to detect people in another room and roughly interpret their movements.

The authors believe that the results of their study show that the model can estimate the precise positioning of multiple objects with performance comparable to image-based approaches using Wi-Fi signals as the only input.

The first line illustrates the hardware configuration. The second and third lines are plots of the amplitude and phase of the Wi-Fi input signal. The fourth line contains the tight pose estimate of the algorithm based solely on the Wi-Fi signal © F. de La Torre et al., 2023

Ethics and privacy

Aside from the technological prowess, the authors make no mention of the ethical issues this raises. Indeed, as noted in the Vice article, the team believes their study should be seen as progress on the right to privacy.

They write in their article: [le nouveau dispositif] protects the privacy of people, and the necessary equipment can be purchased at a reasonable price.” They add: “In fact, most homes in developed countries already have Wi-Fi in the house, and this technology can be scaled up to monitor the well-being of older people or simply detect suspicious behavior at home.”

But they never elaborate on what “suspicious behavior” might include if the technology were to ever enter the mass market. Who will judge the analysis of edited videos? Is this a serious invasion of privacy even when connected via Wi-Fi? All of these questions deserve to be addressed with further research.

While their model is still limited to publicly available training data, the team plans to collect “multi-mock data” and expand their work by predicting the shape of the human body in 3D using Wi-Fi signals.

Drones exploit Wi-Fi loophole

You should know that a research team at the University of Waterloo developed a drone-based device in 2022 that can also use Wi-Fi networks to see through walls.

Dubbed the Wi-Peep, the device can fly close to a building and then use the occupants’ Wi-Fi network to identify and locate all the Wi-Fi-enabled devices inside, from the router to the connected watch, in seconds.

According to the statement, Wi-Peep exploits a flaw that the researchers call “polite Wi-Fi.” Even if the network is password protected, smart devices will automatically respond to contact attempts from any device within range. Wi-Peep sends multiple messages to a device during flight, then measures the response time of each one, allowing the device to be located within a meter.

The researchers warn against extrapolation: “Using technology like this, it is possible to track the movements of security guards inside a bank by tracking the location of their phones or smartwatches. Similarly, a thief can determine the location and type of smart devices in a home, including security cameras, laptops, and smart TVs, in order to find a suitable hack candidate. In addition, drone control of the device means it can be used quickly and remotely with minimal chance of user detection.”

Therefore, the team is urging Wi-Fi chip manufacturers to introduce artificial random fluctuations in device response time, which would make calculations like those used by Wi-Peep highly inaccurate.

archive

Back to top button

Adblock Detected

Please consider supporting us by disabling your ad blocker.