hEYEbrid

Abstract

We introduce hEYEbrid, a calibration-free method for spontaneous and long-term eye gaze tracking, with competitive gaze estimation. It is based on a hybrid concept that combines infrared eye images with corneal imaging. For this, two eye cameras are mounted on a glasses frame. In this way, the pupil can be tracked quickly with high precision. This information is translated into the corneal image, which is used to create a connection to the environment, acting like a scene camera. In a user study with 20 participants, we evaluated our approach against an extended version of the system, called 3C-hEYEbrid, and a state-of-the-art head-mounted Pupil Labs eye tracker. We show that hEYEbrid provides accurate gaze estimation in unconstrained environments and is robust against calibration drift (e.g. caused by taking off and putting on the device). In addition, we present a mobile and wearable implementation of hEYEbrid and 3C-hEYEbrid, that is also usable with a monocular Pupil Labs eye tracker. It connects the head-mounted device to a mobile phone, enabling gaze estimation in real time. hEYEbrid represents a significant step towards pervasive gaze-based interfaces.

Smooth-i

“…It was able to improve a detected inaccuracy of 2.64 • to 1.01 • . Accuracies of 1 • are considered good results [Lander et al 2018], even if greater results are often reported in controlled environments for remote eye trackers [Feit et al 2017;Nyström et al 2013].…”

“…Accuracy is evaluated in degrees of visual angle and computed by the difference in position between two paired points (gaze and target), and comparing it with a desired target accuracy. A predefined accuracy threshold of 1.5 • was set for the evaluation of the method, as remote trackers' average accuracy is often found to be larger than 1 degree [Feit et al 2017;Lander et al 2018].…”

“…The result is computed by taking the median distance from all the gaze points to each target position, and later the mean between all of the targets. An accuracy of 1 • of visual angle is considered good for remote eye trackers [Lander et al 2018].…”

Location Estimation from an Indoor Selfie

“…Nakazawa et al [22] propose a gaze-tracking algorithm based on the CIS using a bendable LED dot matrix system assisted by an infrared camera. Lander et al [23] propose a work for computing 3D gaze from the 2D gaze with the help of infrared cameras and scene cameras. Ohshima et al [24] try to match the CI with the scene pictures in the database by neural network, showing the possible privacy security risk and the prospect of human-machine interaction for CIS.…”