We present a machine learning technique for recognizing discrete gestures and 
estimating continuous 3D hand position for mobile interaction. Our multi-stage 
random forest pipeline jointly classifies hand shapes and regresses metric 
depth of the hand from a single RGB camera. Our technique runs in real time on 
unmodified mobile devices, such as smartphones, smartwatches, and 
smartglasses, complementing existing interaction paradigms with in-air gestures.