Gaze estimation problem tackled through synthetic images
Date
2020Author
Version
Acceso abierto / Sarbide irekia
Type
Contribución a congreso / Biltzarrerako ekarpena
Version
Versión aceptada / Onetsi den bertsioa
Project Identifier
ES/1PE/TIN2016-84388-R
Impact
|
10.1145/3379156.3391368
Abstract
In this paper, we evaluate a synthetic framework to be used in the field of gaze estimation employing deep learning techniques. The lack of sufficient annotated data could be overcome by the utilization of a synthetic evaluation framework as far as it resembles the behavior of a real scenario. In this work, we use U2Eyes synthetic environment employing I2Head datataset as real benchmark for compa ...
[++]
In this paper, we evaluate a synthetic framework to be used in the field of gaze estimation employing deep learning techniques. The lack of sufficient annotated data could be overcome by the utilization of a synthetic evaluation framework as far as it resembles the behavior of a real scenario. In this work, we use U2Eyes synthetic environment employing I2Head datataset as real benchmark for comparison based on alternative training and testing strategies. The results obtained show comparable average behavior between both frameworks although significantly more robust and stable performance is retrieved by the synthetic images. Additionally, the potential of synthetically pretrained models in order to be applied in user's specific calibration strategies is shown with outstanding performances. [--]
Subject
Neural networks,
Datasets gaze estimation
Publisher
Association for Computing Machinery (ACM)
Published in
ETRA'20 Short Papers: ACM Symposium on Eye Tracking Research and Applications, 2020:16
Description
Trabajo presentado al Symposium on Eye Tracking Research and Applications (ETRA ’20 Short Papers). Stuttgart, 2020
Departament
Universidad Pública de Navarra. Departamento de Ingeniería Eléctrica, Electrónica y de Comunicación /
Nafarroako Unibertsitate Publikoa. Ingeniaritza Elektrikoa, Elektronikoa eta Telekomunikazio Ingeniaritza Saila
Publisher version
Sponsorship
The authors would like to acknowledge the Spanish Ministry of Science, Innovation and Universities for their support under Contract TIN2017-84388-R.