This repository shows an implementation for analyzing gaze behavior of a person in a mirror exposure task. This tool allows to define visual bounding boxes derived from measured body data on a mirror image of a person. Together with gaze data from pupil lab eye trackers, it is then possible to analyze gaze behavior. The repository provides the code to the paper: link and includes the sample implementation for the study at that time with the Areas of Interest: head, left hand, right hand, and feet. For more information please read the paper.
- Install all required python packages (see requirements.txt) -> pip install -r requirements.txt
- Start the pipeline by starting the
AoI_creator.pyscript. - See Results in the Results folder afterwards
- Exact documentation will follow. There are some hints in the code where to edit values.
When using this Framework please reference:
@inproceedings{dollinger2022eyetracking,
title = {Analyzing Eye Tracking Data in Mirror Exposure},
author = {Döllinger, Nina and Göttfert, Christopher and Wolf, Erik and Mal, David and Latoschik, Marc Erich and Wienrich, Carolin},
booktitle = {Proceedings of the Conference on Mensch und Computer},
year = {2022},
url = {https://downloads.hci.informatik.uni-wuerzburg.de/2022-muc-eyetracking_in_mirror_exposition-preprint.pdf}
}- Christopher Göttfert (Christopher.Goettfert@uni-wuerzburg.de)
- Nina Döllinger
- Carolin Wienrich