Automatic recognition of touch gestures in the corpus of social touch

Merel Madeleine Jung, Mannes Poel, Ronald Walter Poppe, Dirk K.J. Heylen

    Research output: Contribution to journalArticleAcademicpeer-review

    12 Citations (Scopus)
    103 Downloads (Pure)


    For an artifact such as a robot or a virtual agent to respond appropriately to human social touch behavior, it should be able to automatically detect and recognize touch. This paper describes the data collection of CoST: Corpus of Social Touch, a data set containing 7805 captures of 14 different social touch gestures. All touch gestures were performed in three variants: gentle, normal and rough on a pressure sensor grid wrapped around a mannequin arm. Recognition of these 14 gesture classes using various classifiers yielded accuracies up to 60 %; moreover, gentle gestures proved to be harder to classify than normal and rough gestures. We further investigated how different classifiers, interpersonal differences, gesture confusions and gesture variants affected the recognition accuracy. Finally, we present directions for further research to ensure proper transfer of the touch modality from interpersonal interaction to areas such as human–robot interaction (HRI).
    Original languageEnglish
    Pages (from-to)81-96
    Number of pages16
    JournalJournal on multimodal user interfaces
    Issue number1
    Publication statusPublished - Mar 2017


    • Social Touch
    • EWI-27340
    • IR-102549
    • Touch gesture recognition
    • METIS-319467
    • Touch corpus


    Dive into the research topics of 'Automatic recognition of touch gestures in the corpus of social touch'. Together they form a unique fingerprint.

    Cite this