4 researchers at the University of Washington leverage wireless signals to enable whole-home sensing and recognition of human gestures. It’s called WiSee. Not WiWillSee. WiSee.
Wii and Xbox Kinect have a lot to answer for. The commercialization of gesture based interactivity might be taking a new step forward based no the work of Qifan Pu, Sidhant Gupta, Shyamnath Gollakota, and Shwetak Patel at University of Washington.
The full paper, Whole-Home Gesture Recognition Using Wireless Signals, can be found here.
In summary, WiSee, the system detects the reflection of WiFi signals from a person and assesses the gesture based on relative shifts in the signal. It leverages the property of the Doppler shift, which is the change of a wave as its source moves relative to the observer. Gestures result in very small Doppler shifts, and WiSee detects those shifts and translates them into commands.
For multiple users in a room, each user is identified by a gesture signature, a repetitive gesture, that signals it is their turn to use the system.
The system can detect 9 whole-body movements within 94% accuracy right now in a proof-of-concept lab demonstration unit.
This video is courtesy of the originating article by Michelle Ma.
