Refraction reloaded – Simplifying eye model fitting and refraction modeling
I am happy to announce, that recently my latest computer vision paper was published at ETRA 2019, a major conference in the field of eye tracking. In this collaborative work which was performed at Pupil Labs with Moritz Kassner and Andreas Bulling, we extend on our recent results with respect to eye model fitting and refraction modeling. In particular, we show that eyeball position can be determined from images recorded by a near eye camera in a much more simple and direct fashion as compared to the methods proposed in earlier works. This conceptual simplification entails a vast boost in numerical performance and estimation speed. Secondly, we also develop a method for quantitatively estimating optical effects occuring due to light being bend when leaving or entering the eye due to refraction. On the basis of these results, we devise a method for correcting for these effects in an exisiting gaze estimation algorithm. We were very happy that our work was also selected for an oral presentation at ETRA 2019 in Denver this year.
If you are interested in the original manuscript, you can find a link to it here:
K. Dierkes, M. Kassner, A. Bulling. A fast approach to refraction-aware eye-model fitting and gaze prediction. In ETRA ’19: 2019 Symposium on Eye Tracking Research and Applications, June 25–28, 2019, Denver, CO, USA. ACM, New York, NY, USA.