whole-home gesture recognition整个家的手势识别.pdf

Whole-Home Gesture Recognition Using Wireless Signals Qifan Pu, Sidhant Gupta, Shyamnath Gollakota, and Shwetak Patel University of Washington {qp, sidhant, gshyam, shwetak}@cs.washington.edu Abstract – This paper presents WiSee, a novel gesture home or building. Given these limitations, researchers have recognition system that leverages wireless signals (e.g., Wi- explored ways to move some of the sensing onto the body and Fi) to enable whole-home sensing and recognition of human reduce the need for environmental sensors [8, 14, 13]. How- gestures. Since wireless signals do not require line-of-sight ever, even on-body approaches are limited to what people and can traverse through walls, WiSee can enable whole- are willing to constantly carry or wear, and may be infeasi- home gesture recognition using few wireless sources. Further, ble in many scenarios (e.g., in a shower). it achieves this goal without requiring instrumentation of the This paper presents WiSee, the first whole-home gesture human body with sensing devices. We implement a proof-of- recognition system that requires neither user instrumenta- concept prototype of WiSee using USRP-N210s and evaluate tion nor an infrastructure of cameras. WiSee achieves this it in both an office environment and a two-bedroom apart- by leveraging wireless signals (e.g. Wi-Fi) in an environ- ment. Our results show that WiSee can identify and classify ment. Since these signals do not require line-of-sight and a set of nine gestures with an average accuracy of 94%.