Publication Details

Gaze and Conversational Engagement in Multiparty Video Conversation: An annotation scheme and classification

BEDNAŘÍK Roman, EIVAZI Shahram and HRADIŠ Michal. Gaze and Conversational Engagement in Multiparty Video Conversation: An annotation scheme and classification. In: 4th Workshop on Eye Gaze in Intelligent Human Machine Interaction. Santa Monica: Association for Computing Machinery, 2012, pp. 1-5. ISBN 978-1-4503-1516-6.
Czech title
Pohled a úroveň zapojení do konverzace při telekonferencích: anotační schéma a klasifikace
Type
conference paper
Language
english
Authors
Bednařík Roman (University of Eastern Finland)
Eivazi Shahram (FIT BUT)
Hradiš Michal, Ing., Ph.D. (DCGM FIT BUT)
Keywords

gaze tracking

Abstract

When using a multiparty video mediated system, interacting partic-ipants assume a range of various roles and exhibit behaviors accord-ing to how engaged in the communication they are. In this paper we
focus on estimation of conversational engagement fromgaze signal.
In particular, we present an annotation scheme for conversational
engagement, a statistical analysis of gaze data across varying levels
of engagement, and we classify vectors of computed eye tracking
measures. The results show that in 74% of cases the level of en-gagement can be correctly classified.

Published
2012
Pages
1-5
Proceedings
4th Workshop on Eye Gaze in Intelligent Human Machine Interaction
Conference
14th ACM International Conference on Multimodal Interaction, Santa Monica, US
ISBN
978-1-4503-1516-6
Publisher
Association for Computing Machinery
Place
Santa Monica, US
DOI
10.1145/2401836.2401846
BibTeX 
@INPROCEEDINGS{FITPUB10163,
   author = "Roman Bedna\v{r}\'{i}k and Shahram Eivazi and Michal Hradi\v{s}",
   title = "Gaze and Conversational Engagement in Multiparty Video Conversation: An annotation scheme and classification",
   pages = "1--5",
   booktitle = "4th Workshop on Eye Gaze in Intelligent Human Machine Interaction",
   year = 2012,
   location = "Santa Monica, US",
   publisher = "Association for Computing Machinery",
   ISBN = "978-1-4503-1516-6",
   doi = "10.1145/2401836.2401846",
   language = "english",
   url = "https://www.fit.vut.cz/research/publication/10163"
}
Back to top