Using Corneal Imaging for Measuring a Human's Visual Attention

Christian Lander; Felix Kosmalla; Frederik Wiehr; Sven Gehring

In: Proceedings of the 2017 ACM International Joint Conference on Pervasive and Ubiquitous Computing and Proceedings of the 2017 ACM International Symposium on Wearable Computers. International Joint Conference on Pervasive and Ubiquitous Computing (UbiComp-17), September 11-15, Maui, Hawaii, USA, Pages 947-952, UbiComp '17, ISBN 978-1-4503-5190-4, ACM, New York, NY, USA, 2017.


The human's visual attention focus usually reflects the activity and environment engaged in or simply his context in the most coherent way. The cornea, which is encasing the iris and pupil, is protected by tear fluid, and thus is a highly reflective surface. Our eyes show a reflection of what we see in our current context. We present an attempt of using corneal imaging to extract contextual information - including objects in the field of view - and the user's attention focus. Our system uses a head-mounted eye-camera, for capturing corneal reflections, connected to a RaspberryPi, to record a humans' current view and estimate his focus of attention. We conducted a 2-day experiment, where we collected data in un-instrumented real-world settings. Based on the analysis of the recordings we illustrate what kind of information can be extracted out of the corneal image reflections and outline the possibilities for in-situ notifications.


Weitere Links

3123024.3124563.pdf (pdf, 1 MB )

Deutsches Forschungszentrum für Künstliche Intelligenz
German Research Center for Artificial Intelligence