RSTG 2015 - Recognition of Social Touch Gestures Challenge 2015

Touch is one of the important non-verbal forms of social interaction, where it is used to communicate emotions and social messages. Automatic recognition of social touch is necessary to apply interaction in the tactile modality from to areas such as Human-Robot Interaction (HRI). If a robot can understand social touch behavior, the robot can respond accordingly - resulting in richer and more natural interaction.
In this challenge the focus is on the recognition of different hand touch gestures for which two data sets will be made available, each with labeled pressure/ location data collected from similar matrix-type sensor grids under conditions reflecting different application orientations. (1) CoST: Corpus of Social Touch (Jung et al., 2014) includes 14 relevant touch gestures such as stroke, poke and hit performed in three variations: normal, gentle and rough. Challenge: gentle (2601 captures) and normal (2602 captures) variations of the 14 gestures used. (2) HAART: The Human-Animal Affective Robot Touch gesture set (Flagg and MacLean, 2013) includes 7 touch gestures which humans use to communicate emotion in animal interaction (Yohanan and MacLean, 2012), with a range of covering, substrate and curvature. Challenge: one covering, substrate and distortion used.
Participants can choose to work on one of the data sets or on both. The purpose of this challenge is to develop relevant features and classification methods for recognizing social touch gestures. Participants will share their innovative findings at ICMI 2015.
Challenge homepage: http://www.utwente.nl/touch-challenge-2015