Online attention for interpretable conflict estimation in political debates

Ruben Vereecken, Stavros Petridis, Yiannis Panagakis, Maja Pantic

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

    1 Citation (Scopus)

    Abstract

    Conflict arises naturally in dyadic interactions when involved individuals act on incompatible goals, interests, or actions. In this paper, the problem of conflict intensity estimation from audiovisual recordings is addressed. To this end, we propose an online attention-based neural network in order to learn a mapping from a sequence of audiovisual features to time-series describing conflict intensity. The proposed method is evaluated by conducting experiments in conflict intensity estimation by employing the CONFER dataset. Experimental results indicate the superiority of the proposed model compared to the state of the art. Furthermore, we demonstrate that by incorporating sparsity in the model, the origin of conflict can be traced back to specific key frames facilitating the interpretation of conflict escalation.

    Original languageEnglish
    Title of host publication13th IEEE International Conference on Automatic Face and Gesture Recognition, FG 2018
    PublisherIEEE
    Pages389-393
    Number of pages5
    ISBN (Electronic)9781538623350
    DOIs
    Publication statusPublished - 5 Jun 2018
    Event13th IEEE International Conference on Automatic Face and Gesture Recognition, FG 2018: First Workshop on Large-scale Emotion Recognition and Analysis - Xi'an, China
    Duration: 15 May 201819 May 2018
    Conference number: 13
    https://fg2018.cse.sc.edu/

    Conference

    Conference13th IEEE International Conference on Automatic Face and Gesture Recognition, FG 2018
    Abbreviated titleFG 2018
    CountryChina
    CityXi'an
    Period15/05/1819/05/18
    Internet address

    Keywords

    • Attention
    • CONFER
    • Conflict
    • Debate
    • Estimation
    • Network
    • Political

    Fingerprint

    Dive into the research topics of 'Online attention for interpretable conflict estimation in political debates'. Together they form a unique fingerprint.

    Cite this