New technology utilizes Wi-Fi signals to interpret user gestures into commands from anywhere inside a home or office.
Scientists at the University of Washington have developed a method of recognizing a user’s gestures regardless of their physical location within a building, using nothing more than existing Wi-Fi signals and a modified receiver device.
Known as WiSee, the concept is similar to Mirosoft’s Xbox Kinect. Unlike the commercial gaming system accessory, however, which uses cameras to recognize gestures and requires line-of-sight to function, the WiSee technology relies on an adapted Wi-Fi router and existing Wi-Fi signals to detect and interpret a user’s commands from any physical location within a home or office.
“This is repurposing wireless signals that already exist in new ways,” lead researcher Shyam Gollakota, a UW assistant professor of computer science and engineering, said. “You can actually use wireless for gesture recognition without needing to deploy more sensors.”
According to UW researchers, the technology works by identifying small changes in the frequency of the wireless signal caused by physical movement. “These frequency changes are very small—only several hertz—when compared with Wi-Fi signals that have a 20 megahertz bandwidth and operate at 5 gigahertz.”
Using a “smart” receiver device and an algorithm developed to detect and record these frequency changes, researchers identified certain patterns of changes—known as the Doppler frequency shift—and linked them with corresponding gestures. The new technology is capable of identifying nine different full-body gestures, including pushing, pulling punching and full-body bowling.
During tests completed in a two-bedroom apartment and office environment, WiSee accurately classified about 94 percent of the 900 gestures performed. Multiple antennas equipped to the receiver enable as many as five people to simultaneously move throughout the building without confusing the device.
The technology also accounts for gaps in wireless signals when wireless devices are not transmitting.
“This is the first whole-home gesture recognition system that works without either requiring instrumentation of the user with sensors or deploying cameras in every room,” Qifan Pu, a project collaborator and visiting student at the UW, said.
The technology could one day be used to realize the concept of a smart home, enabling users to turn on the coffeemaker from the bedroom, change the song playing on the living room stereo from the kitchen and pause a movie during a bathroom break with a simple hand movement.
The research team plans to continue their research by investigating the possibility of using WiSee to control multiple devices at once.
