CoST includes 14 touch gestures: grab, hit, massage, pat, pinch, poke, press, rub, scratch, slap, squeeze, stroke, tap and tickle. The gestures are registered by a 8 by 8 pressure sensor grid which is wrapped around a mannequin arm. The corpus consists of the data of 31 subjects who performed 14 touch gestures in 3 variations (normal, gentle and rough). For the challenge the data of the gentle gesture variation (2601 captures) and the normal gesture variations (2602 captures) will be made available. The complete dataset will be publicly available after the challenge. Pressure values of the 64 channels range from 0 to 1023 (i.e. 10 bits) and are sampled at 135 Hz. Labels consist of gesture type, variation, and subject number. Segmented data will be provided in CSV and MATLAB format. Readers can refer to [1,2] for more detailed information about CoST. The full data set is available online .
 M. M. Jung, R. Poppe, M. Poel, and D. K. J. Heylen (2014). Touching the void – introducing CoST: Corpus of Social Touch. In Proceedings of the International Conference on Multimodal Interaction (ICMI), (Istanbul, Turkey), pages 120–127.
 M. M. Jung, M. Poel, R. Poppe, and D. K. J. Heylen (2016). Automatic recognition of touch gestures in the corpus of social touch. J Multimodal User Interfaces. doi:10.1007/s12193-016-0232-9
 M. M. Jung, M. Poel, R. Poppe, and D. K. J. Heylen (2016). Corpus of Social Touch (CoST). University of Twente. http://dx.doi.org/10.4121/uuid:5ef62345-3b3e-479c-8e1d-c922748c9b29