RT-GENE: Real-time eye gaze estimation in natural environments

File Description SizeFormat 
Fischer_ECCV2018_RT-GENE_stamped2.pdfAccepted version4.96 MBAdobe PDFDownload
Title: RT-GENE: Real-time eye gaze estimation in natural environments
Author(s): Fischer, T
Chang, HJ
Demiris, Y
Item Type: Conference Paper
Abstract: In this work, we consider the problem of robust gaze estimation in natural environments. Large camera-to-subject distances and high variations in head pose and eye gaze angles are common in such environments. This leads to two main shortfalls in state-of-the-art methods for gaze estimation: hindered ground truth gaze annotation and diminished gaze estimation accuracy as image resolution decreases with distance. We first record a novel dataset of varied gaze and head pose images in a natural environment, addressing the issue of ground truth annotation by measuring head pose using a motion capture system and eye gaze using mobile eyetracking glasses. We apply semantic image inpainting to the area covered by the glasses to bridge the gap between training and testing images by removing the obtrusiveness of the glasses. We also present a new real-time algorithm involving appearance-based deep convolutional neural networks with increased capacity to cope with the diverse images in the new dataset. Experiments with this network architecture are conducted on a number of diverse eye-gaze datasets including our own, and in cross dataset evaluations. We demonstrate state-of-the-art performance in terms of estimation accuracy in all experiments, and the architecture performs well even on lower resolution images.
Publication Date: 6-Oct-2018
Date of Acceptance: 3-Jul-2018
URI: http://hdl.handle.net/10044/1/62579
DOI: https://dx.doi.org/10.1007/978-3-030-01249-6_21
ISSN: 0302-9743
Publisher: Springer Verlag
Start Page: 334
End Page: 352
Journal / Book Title: Lecture Notes in Computer Science
Volume: 11214
Copyright Statement: © Springer Nature Switzerland AG 2018. The final publication is available at Springer via https://link.springer.com/chapter/10.1007/978-3-030-01249-6_21
Sponsor/Funder: Commission of the European Communities
Samsung Electronics Co Ltd
Funder's Grant Number: 643783
Conference Name: European Conference on Computer Vision
Keywords: Gaze estimation
Gaze dataset
Convolutional Neural Network
Semantic inpainting
Eyetracking glasses
08 Information And Computing Sciences
Artificial Intelligence & Image Processing
Start Date: 2018-09-08
Finish Date: 2018-09-14
Conference Place: Munich, Germany
Open Access location: http://openaccess.thecvf.com/content_ECCV_2018/html/Tobias_Fischer_RT-GENE_Real-Time_Eye_ECCV_2018_paper.html
Appears in Collections:Faculty of Engineering
Electrical and Electronic Engineering

Items in Spiral are protected by copyright, with all rights reserved, unless otherwise indicated.

Creative Commons