Automatic Segmentation of Spontaneous Data using Dimensional Labels from Multiple Coders

Mihalis A. Nicolaou, Hatice Gunes, Maja Pantic

    Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

    27 Downloads (Pure)

    Abstract

    This paper focuses on automatic segmentation of spontaneous data using continuous dimensional labels from multiple coders. It introduces efficient algorithms to the aim of (i) producing ground-truth by maximizing inter-coder agreement, (ii) eliciting the frames or samples that capture the transition to and from an emotional state, and (iii) automatic segmentation of spontaneous audio-visual data to be used by machine learning techniques that cannot handle unsegmented sequences. As a proof of concept, the algorithms introduced are tested using data annotated in arousal and valence space. However, they can be straightforwardly applied to data annotated in other continuous emotional spaces, such as power and expectation.
    Original languageUndefined
    Title of host publicationWorkshop on Multimodal Corpora: Advances in Capturing, Coding and Analyzing Multimodality
    EditorsMichale Kipp, Jean-Claude Martin, Patrizia Paggio, Dirk K.J. Heylen
    Place of PublicationSaarbrucken
    PublisherGerman Research Center for AI (DFKI)
    Pages43-48
    Number of pages6
    ISBN (Print)not assigned
    Publication statusPublished - 18 May 2010
    EventWorkshop on Multimodal Corpora: Advances in Capturing, Coding and Analyzing Multimodality - Valletta, Malta
    Duration: 18 May 201018 May 2010

    Publication series

    Name
    PublisherGerman Research Center for AI (DFKI)

    Workshop

    WorkshopWorkshop on Multimodal Corpora: Advances in Capturing, Coding and Analyzing Multimodality
    Period18/05/1018/05/10
    Other18 May 2010

    Keywords

    • IR-75993
    • METIS-276354
    • EC Grant Agreement nr.: FP7/211486
    • HMI-MI: MULTIMODAL INTERACTIONS
    • EWI-19537

    Cite this