µPoly: a First Microgesture Toolkit EICS004
Adrien Chaffangeon Caillet, Aurélien Conil, Vincent Lambert, Charles Bailly, Alix Goguey, Laurence NigayWith the rapid development of microgesture research, numerous recognition devices with different sensors (e.g. force-sensitive, IMU), form factors (e.g. glove, ring) and recognition algorithms (e.g. ad-hoc, random forest) have become available. With this increasing number and diversity of recognition devices, it is important to be able to easily switch between devices and/or modify the set of recognized microgestures. To address these challenges, we propose µPoly, a first microgesture toolkit based on the µGlyph microgesture notation. µPoly takes as input the µGlyph descriptions of elementary microgestures. These inputs are combined to recognize complex microgestures described using µGlyph. µPoly therefore offers a standard for microgesture events. The design of µPoly is based on properties taken from the toolkit literature. To highlight these properties: we present (1) how practitioners are using the toolkit in four scenarios, and report on user testing and (2) how applications are being developed with µPoly.