3D eye model-based gaze estimation from a depth sensor

Xiaolong Zhou, Haibin Cai, Zhanpeng Shao, Hui Yu, Honghai Liu

Research output: Chapter in Book/Report/Conference proceedingConference contribution

542 Downloads (Pure)


In this paper, we address the 3D eye gaze estimation problem using a low-cost, simple-setup, and nonintrusive consumer depth sensor (Kinect sensor). We present an effective and accurate method based on 3D eye model to estimate the point of gaze of a subject with the tolerance of free head movement. To determine the parameters involved in the proposed eye model, we propose i) an improved convolution-based means of gradients iris center localization method to accurately and efficiently locate the iris center in 3D space; ii) a geometric constraints-based method to estimate the eyeball center under the constraints that all the iris center points are distributed on a sphere originated from the eyeball center and the sizes of two eyeballs of a subject are identical; iii) an effective Kappa angle calculation method based on the fact that the visual axes of both eyes intersect at a same point with the screen plane. The final point of gaze is calculated by using the estimated eye model parameters. We experimentally evaluate our gaze estimation method on five subjects. The experimental results show the good performance of the proposed method with an average estimation accuracy of 3.78, which outperforms several state-of-the-arts.
Original languageEnglish
Title of host publicationProceedings of the 2016 IEEE International Conference on Robotics and Biomimetics
PublisherInstitute of Electrical and Electronics Engineers Inc.
Number of pages6
ISBN (Electronic)978-1509043644
ISBN (Print)978-1509043651
Publication statusPublished - 2 Mar 2017
Event2016 IEEE International Conference on Robotics and Biomimetics - Qingdao, China
Duration: 3 Dec 20167 Dec 2016


Conference2016 IEEE International Conference on Robotics and Biomimetics
Abbreviated titleROBIO 2016


Dive into the research topics of '3D eye model-based gaze estimation from a depth sensor'. Together they form a unique fingerprint.

Cite this