A Semi-automated System for Accurate Gaze Coding in Natural Dyadic Interactions. Funes Mora, K. A., Nguyen, L., Gatica-Perez, D., & Odobez, J. In Proceedings of the 15th ACM on International Conference on Multimodal Interaction, of ICMI '13, pages 87--90, New York, NY, USA, 2013. ACM.
A Semi-automated System for Accurate Gaze Coding in Natural Dyadic Interactions [link]Paper  doi  abstract   bibtex   
In this paper we propose a system capable of accurately coding gazing events in natural dyadic interactions. Contrary to previous works, our approach exploits the actual continuous gaze direction of a participant by leveraging on remote RGB-D sensors and a head pose-independent gaze estimation method. Our contributions are: i) we propose a system setup built from low-cost sensors and a technique to easily calibrate these sensors in a room with minimal assumptions; ii) we propose a method which, provided short manual annotations, can automatically detect gazing events in the rest of the sequence; iii) we demonstrate on substantially long, natural dyadic data that high accuracy can be obtained, showing the potential of our system. Our approach is non-invasive and does not require collaboration from the interactors. These characteristics are highly valuable in psychology and sociology research.
@inproceedings{funes_mora_semi-automated_2013,
	address = {New York, NY, USA},
	series = {{ICMI} '13},
	title = {A {Semi}-automated {System} for {Accurate} {Gaze} {Coding} in {Natural} {Dyadic} {Interactions}},
	isbn = {978-1-4503-2129-7},
	url = {http://doi.acm.org/10.1145/2522848.2522884},
	doi = {10.1145/2522848.2522884},
	abstract = {In this paper we propose a system capable of accurately coding gazing events in natural dyadic interactions. Contrary to previous works, our approach exploits the actual continuous gaze direction of a participant by leveraging on remote RGB-D sensors and a head pose-independent gaze estimation method. Our contributions are: i) we propose a system setup built from low-cost sensors and a technique to easily calibrate these sensors in a room with minimal assumptions; ii) we propose a method which, provided short manual annotations, can automatically detect gazing events in the rest of the sequence; iii) we demonstrate on substantially long, natural dyadic data that high accuracy can be obtained, showing the potential of our system. Our approach is non-invasive and does not require collaboration from the interactors. These characteristics are highly valuable in psychology and sociology research.},
	urldate = {2014-06-05TZ},
	booktitle = {Proceedings of the 15th {ACM} on {International} {Conference} on {Multimodal} {Interaction}},
	publisher = {ACM},
	author = {Funes Mora, Kenneth A. and Nguyen, Laurent and Gatica-Perez, Daniel and Odobez, Jean-Marc},
	year = {2013},
	pages = {87--90}
}

Downloads: 0