Breaking the status quo: Improving 3D gesture recognition with spatially convenient input devices