Neural Conditional Ordinal Random Fields for Agreement Level Estimation

Nemanja Rakicevic, Ognjen Rudovic, Stavros Petridis

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

    1 Citation (Scopus)
    21 Downloads (Pure)


    We present a novel approach to automated estimation of agreement intensity levels from facial images. To this end, we employ the MAHNOB Mimicry database of subjects recorded during dyadic interactions, where the facial images are annotated in terms of agreement intensity levels using the Likert scale (strong disagreement, disagreement, neutral, agreement and strong agreement). Dynamic modelling of the agreement levels is accomplished by means of a Conditional Ordinal Random Field model. Specifically, we propose a novel Neural Conditional Ordinal Random Field model that performs non-linear feature extraction from face images using the notion of Neural Networks, while also modelling temporal and ordinal relationships between the agreement levels. We show in our experiments that the proposed approach outperforms existing methods for modelling of sequential data. The preliminary results obtained on five subjects demonstrate that the intensity of agreement can successfully be estimated from facial images (39% F1 score) using the proposed method.
    Original languageUndefined
    Title of host publicationProceedings of the International Conference on Affective Computing and Intelligent Interaction (ACII 2015)
    EditorsN. Rakicevic, O. Rudovic, S. Petrids, Maja Pantic
    Place of PublicationUSA
    PublisherIEEE Communications Society
    Number of pages6
    ISBN (Print)978-1-4799-9953-8
    Publication statusPublished - Sep 2015
    Event6th International Conference on Affective Computing and Intelligent Interaction, ACII 2015 - Xi'an, China
    Duration: 21 Sep 201524 Sep 2015
    Conference number: 6

    Publication series

    PublisherIEEE Communications Society


    Conference6th International Conference on Affective Computing and Intelligent Interaction, ACII 2015
    Abbreviated titleACII


    • HMI-HF: Human Factors
    • EWI-25800
    • conditional ordinal random fields
    • EC Grant Agreement nr.: FP7/611153
    • METIS-316020
    • Neural Networks
    • IR-99504
    • Agreement analysis
    • EC Grant Agreement nr.: FP7/2007-2013

    Cite this