Science

Wi-Fi tech identifies people through walls – by their walk

Wi-Fi tech identifies people t...
It is suggested that the system could be used to tell if a person hiding in a house was actually a bank robber, who was earlier recorded by a bank's security cameras
It is suggested that the system could be used to tell if a person hiding in a house was actually a bank robber, who was earlier recorded by a bank's security cameras
View 1 Image
It is suggested that the system could be used to tell if a person hiding in a house was actually a bank robber, who was earlier recorded by a bank's security cameras
1/1
It is suggested that the system could be used to tell if a person hiding in a house was actually a bank robber, who was earlier recorded by a bank's security cameras

Imagine if a robber was recorded on a bank's security-camera system, and then perhaps hid in a certain house. Utilizing a new technique, ordinary wall-penetrating Wi-Fi signals could reportedly now be used to determine if the person in that house was indeed the robber.

Developed by a team at the University of California-Santa Barbara, the XModal-ID ("cross-modal ID") system begins by using previously-recorded video footage of a person to ascertain the unique pattern of their walking gait. A 3D computer model based on that gait pattern is then used to determine the telltale manner in which that person would disrupt a wireless signal as they walked through it.

If it's suspected that the individual may be inside a specific room or building, two devices – a Wi-Fi transmitter and a Wi-Fi receiver – are set up outside that structure's walls. As the wireless signal passes from the one device to the other, it travels through the structure, getting partially absorbed by objects (such as people) within.

By analyzing the fluctuations in the strength of the signal as it reaches the receiver, it's possible to determine if the disruptions are being caused by the person in question, as they walk around. In tests that involved 1,488 combinations of Wi-Fi data and gait-pattern video – drawn from a pool of eight test subjects walking in three different behind-wall settings – the system proved to have an overall accuracy rate of 84 percent at identifying the individuals. That rate should rise as the technology is developed further.

"Our proposed approach makes it possible to determine if the person behind the wall is the same as the one in video footage, using only a pair of off-the-shelf Wi-Fi transceivers outside," says study leader Prof. Yasmin Mostofi. "This approach utilizes only received power measurements of a Wi-Fi link. It does not need any prior Wi-Fi or video training data of the person to be identified. It also does not need any knowledge of the operation area."

XModal-ID could conceivably also find use in non-law-enforcement applications, such as the identification of family members within a smart home.

Source: UC Santa Barbara

0 comments
There are no comments. Be the first!