Touchi: Online Touch User Identification through Wrist-worn Inertial Measurement Units
This project is a collaboration by Ahmed Kharrufa, Nils Hammerla, Phil Heslop and Thomas Ploetz. User identification for multi-touch surfaces is a highly desirable feature that is not inherently supported by existing touch sensing technologies. We present TOUCHI, a novel approach for touch user identification through the use of wrist-worn inertial movement units (IMU). Our approach for user identification is based on identifying the wrist-worn IMU signal that shows the highest similarity to the touch data obtained from a touch sensitive surface. We also introduce a novel 6- axis conceptual model for user identification in multi- touch scenarios for characterizing the implications of different application requirements with regards to user identification, and with which we can contrast the strengths and weaknesses of TOUCHI compared to other user identification approaches.