The Malta Independent on Sunday
University of Malta’s WildEye project seeks to capture the eye-gaze tracking market
It is often said that the eyes are windows to the soul, and indeed research indicates that the human gazing behaviour is often indicative of underlying cognitive and affective processes. This fundamental human behaviour serves essential functions in our lives; we use it to pick up relevant visual information to perform a task, or to infer information about a person or situation; we establish eye contact in order to communicate and create socio-emotional connections; and we indicate wishes, preferences and attention by directing our gaze to a target of interest. This wealth of information contained within the eye movements has been receiving increasing interest for over a century, leading to the development of different eye-gaze tracking techniques that seek to capture and measure the human eye-gaze. Over the years, eyegaze tracking technology has been applied to a multitude of domains outside the clinical laboratory, such as in human-computer interaction, automotive engineering for car driver assistance and monitoring, humanrobot interaction, marketing and advertising research, sports performance and research, visual perception of art, and more recently human authentication.
A market research of eye-gaze tracking technologies reveals that the global eye tracking market was estimated to be worth over US$183 million in 2014 and is forecasted to reach a billion US dollars by the year 2020, with human-computer interaction being forecasted as one of the key applications of eye-gaze tracking for the next few years. This is evidenced by a growing number of high-technology companies that develop eye-gaze tracking products, leading to vast technological improvements over the years and the availability of a host of low to high-end devices on the market. This growing interest has been heightened further with the recent acquisition of three of the major market players by Google, Apple and Facebook. The next technological leap for eye-gaze tracking is the seamless integration of this technology into consumer products, making this a ubiquitous technology that is available to everyone.
Unobtrusive eye-gaze tracking devices that may be presently purchased off the market come in the form of a dedicated hardware module that houses a set of infra-red (IR) illuminators and IR cameras, and typically connects to a computer via USB. The active tracking device exploits the physiological properties of the face and eyes by projecting IR illumination that creates glints, which reflect off the lens and cornea, or otherwise causes the pupil to appear bright in the im- ages captured by the IR cameras. Detection of the bright glints or pupil in the captured images is facilitated by their stark contrast to the surrounding skin, consequently facilitating the computation of gaze from the relative positioning of the glints and pupil. Nonetheless, the effectiveness of active eye-gaze trackers tends to be hampered by interfering IR illumination from the surrounding environment, which affects the image quality of the pupil size and brightness. The allowable head movement is often restricted by the narrow field-of-view of the cameras, and hence the user is required to move within a confined volume in space to be tracked successfully. These issues are further exacerbated by the need for dedicated hardware, which eliminates the possibility of seamlessly integrating this technology into the devices that we already use every day without requiring hardware modification.
In the absence of any dependence upon specialised IR hardware, the alternative use of a webcam for eye-gaze tracking may potentially offer a better and more portable solution to capture the natural visual behaviour in less constrained scenarios. The vast improvements in imaging hardware technology are increasingly permitting a widespread availability of digital cameras that are small in size, and feature higher image resolution and dynamic range at an affordable cost. Furthermore, their expansive field-of-view widens the volume in which the user is allowed to move, while their ready availability on off-theshelf notebook computers and mobile devices eliminates the need for hardware modification, hence retaining the technology at an affordable price. These characteristics of passive eyegaze tracking contribute towards the vision of making eye-gaze tracking a ubiquitous technology that is accessible to everyone, everywhere. Nonetheless, this endeavour does not come without its own challenges. Indeed, real-life environments may be characterised by varying illumination conditions, larger head movement requires that the head pose is tracked ro- bustly and compensated for in the estimation of gaze, while not all settings may permit the user to perform calibration, especially if the tracking device is situated in a public space. An overview of existing methods for passive eye-gaze tracking reveals that we are not yet equipped to cater for the emerging challenges of less constrained, real-life scenarios robustly.
Over the past year, the Department of Systems and Control Engineering, at the University of Malta, has teamed up with Seasus Ltd in a project entitled WildEye, funded by the Malta Council for Science and Technology (MCST) under the FUSION R&I Technology Development Programme 2016. Building upon the extensive research experience on eye-gaze tracking that the Department has garnered over the past decade, the goal of the WildEye project is to address the challenges associated with passive eye-gaze tracking in daily life conditions. Several of the research challenges that the WildEye project is addressing include the robust estimation of head movement under non-rigid face deformations, tracking of the eyegaze by joint head and eye pose estimation, and the reduction of user calibration for more natural user interaction. Seasus Ltd are in the process of designing and developing an eye-controlled user interface that will enable the user to drive computer applications using the eye movements alone. Through their joint effort during the WildEye project, the Department of Systems and Control Engineering together with Seasus Ltd aspire to produce a prototype eye-gaze tracking platform that is closer to market. Encouraged by the results obtained so far, the team working on WildEye is motivated to make eye-gaze tracking a ubiquitous technology that is accessible to everyone.
The team working on WildEye is composed of Professor Kenneth P. Camilleri, the Project Coordinator, and Dr Stefania Cristina from the University of Malta, and Mr Kenneth Bone, Mr David Vella and Mr Panos Brikalidis from Seasus Ltd. Interested individuals are invited to visit the project’s Facebook page, www.facebook.com/wildeyemlt, for more information or to leave feedback.