CUBEFLOW

The app CubeFlow provides the user with a fluid experimental modeling experience. The user controls the app with his fingers and hand gestures. Intel Perceptual Computing SDK is used to provide the input pipeline (PXCMPipeline) for gesture recognition (PXCMPipeline.QueryGesture) and to obtain finger positions (PXCMGesture.GeoNodes) in world space (PXCMGesture.GeoNode.positionWorld). The interaction model of CubeFlow is based on two modes, the "recording mode" and the "navigation mode" which can be toggled from the UI or by thumb up/down gestures. If the recording is running, the app is in "recording mode". Incoming finger position of the left and right hand are used to generate volumetric shapes. Each recognized finger of each hand draws a circular shape into the volume. The user can vary the amount of fingers which are used for modeling and can rotate and move them around to actually create the model. As long as the app runs the recording, the inputs are continuously processed into a volumetric 3d shape over time. The closer the fingers are to the camera, the bigger the input shapes scale. If the recording is paused, the app is in "navigation mode" which helps to examine the created model. The user can rotate the model around two axis by moving one or two hands left, right, up and down. Your hand's movements are used to control the rotation when 3 or more fingers are recognized ("open hand"). Otherwise if 2 or less fingers are recognized, no hand movement is used for rotation ("closed hand"), so that the user can reposition his hand for the next navigation action. If the hands move closer to the camera, the created model can be zoomed out. Moving the hands away from the camera the view zooms in.