Gesture Variation Follower (GVF) is a cross-platform, open-source C++ library for realtime gesture recognition and variations estimation. It is has been designed to support human-computer interactions (HCI) and belongs to a kind of technology to which, internally at RAPIDMIX, we refer to as realtime machine learning for motion recognition and classification.
This library provides methods to easily learn a gesture vocabulary, recognize a gesture as soon as it is performed, and estimate its variations (e.g. in scale, orientation, or dynamic). It allows mixing discrete and continuous commands, and integrating with creative applications such as controlling sounds and visuals, and its prototype has been demonstrated in operational environments.
This library is open source, licensed under LGPL v3 by IRCAM and Goldsmiths. It is provided with an openFrameworks add-on (ofxGVF) and also used to build Max/MSP and PureData objects, which are available online. Installation and API are detailed on the Github repo webpage. For any further information see below.