Gesture
recognition is becoming increasingly important in human-PC communication (HCl).

Gesture based interaction fills in as an advantageous and natural means for
clients to collaborate with computers. Gestures made with fingers is especially
significant to cooperate with mobile and wearable gadgets and to perform finger
control in smart home and gaming. For example, Google’s Soli is the recently
developed radar chip 1 for the wearables to detect finger gestures. Existing gesture
detection approaches depend on committed sensors worn by the subject 2 or
cameras installed in the system3. These frameworks either require heavy installation
effort or high cost. Recently, radio frequency (RF) based gesture detection 4,
5, 6 have drawn extensive attention as they don’t need clients to wear any
physical sensors. These frameworks however just give coarse-grained gesture
detection such as body exercises 6, 4 or hand motion 5.

Our system
demonstrates that WiFi can be used for finger gesture detection which is both
effectively deployable and low in cost.