Play with me - Measuring a child's engagement in a social interaction

Shyam RAJAGOPALAN, Ramana ORUGANTI, Roland GOECKE, Agata Rozga

Research output: A Conference proceeding or a Chapter in BookConference contribution

10 Citations (Scopus)

Abstract

Due to the challenges in automatically observing child behaviour in a social interaction, an automatic extraction of high-level features, such as head poses and hand gestures, is difficult and noisy, leading to an inaccurate model. Hence, the feasibility of using easily obtainable low-level optical flow based features is investigated in this work. A comparative study involving high-level features, baseline annotations of multiple modalities and the low-level features is carried out. Optical flow based hidden structure learning of behaviours is strongly discriminatory in predicting a child’s engagement level in a social interaction. A two-stage approach of discovering the hidden structures using Hidden Conditional Random Fields, followed by learning an SVM-based model on the hidden state marginals is proposed. This is validated by conducting experiments on the Multimodal Dyadic Behaviour Dataset and the results indicate a state of the art classification performance. The insights drawn from this study indicate the robustness of the low-level feature approach towards engagement behaviour
modelling and can be a good substitute in the absence of accurate high-level features.
Original languageEnglish
Title of host publication2015 11th IEEE International conference and workshop on Automatic face and gesture recognition (FG 2015)
EditorsKevin Bowyer, Ales Leonardo, Jeff Cohn
Place of PublicationLjubljana, Slovenia
PublisherIEEE
Pages1-8
Number of pages8
Volume1
ISBN (Electronic)9781479960262
DOIs
Publication statusPublished - 4 May 2015
Event11th IEEE International conference and workshop on Automatic face and gesture recognition 2015 - Ljubljana, Ljubljana, Slovenia
Duration: 4 May 20158 May 2015
http://www.fg2015.org/ (Conference website)

Publication series

Name2015 11th IEEE International Conference and Workshops on Automatic Face and Gesture Recognition, FG 2015

Conference

Conference11th IEEE International conference and workshop on Automatic face and gesture recognition 2015
CountrySlovenia
CityLjubljana
Period4/05/158/05/15
OtherThe IEEE conference series on Automatic Face and Gesture Recognition is the premier international forum for research in image and video-based face, gesture, and body movement recognition. Its broad scope includes: advances in fundamental computer vision, pattern recognition and computer graphics; machine learning techniques relevant to face, gesture, and body motion; new algorithms and applications. The conference presents research that advances the state-of-the-art in these and related areas, leading to new capabilities in various application domains
Internet address

Fingerprint Dive into the research topics of 'Play with me - Measuring a child's engagement in a social interaction'. Together they form a unique fingerprint.

  • Cite this

    RAJAGOPALAN, S., ORUGANTI, R., GOECKE, R., & Rozga, A. (2015). Play with me - Measuring a child's engagement in a social interaction. In K. Bowyer, A. Leonardo, & J. Cohn (Eds.), 2015 11th IEEE International conference and workshop on Automatic face and gesture recognition (FG 2015) (Vol. 1, pp. 1-8). (2015 11th IEEE International Conference and Workshops on Automatic Face and Gesture Recognition, FG 2015). IEEE. https://doi.org/10.1109/fg.2015.7163129