The 2024 ACM Symposium on Eye Tracking Research and Applications (ETRA) was held in Glasgow, Scotland, from June 4th to June 7th. Attendees had the opportunity to engage in discussions and attend presentations on cutting-edge research shaping the future of eye-tracking technology and its applications.
Parvin Emami from University of Luxembourg presented the research paper “Impact of Design Decisions in Scanpath Modeling” which was co-authored with Professor Luis A. Leiva and colleagues Yue Jiang and Zixin Guo from Aalto University. The paper underscores the importance of design parameters that are not learnable from data when developing scanpath modeling.
Scanpath modeling aims to predict the sequence of eye movements of each user. Our paper specifically focused on Graphical User Interface (GUI) applications. We showed that even minor variations in design parameters in scanpath modeling, such as image size, inhibition-of-return, and masking radius, can significantly impact the performance of the predictive models. We introduced a set of design parameters for scanpath models that generalize to many scanpath models and even beyond GUI-only scenarios.
The insights from this paper are key to SYMBIOTIK as they help improving the reliability of eye-tracking methodology, enabling more reliable prediction of user visual attention to enhance the adaptation of visualizations, which aligns with the project’s ultimate goal of creating a symbiotic human-AI cooperation.
References
- Parvin Emami, Yue Jiang, Zixin Guo, Luis A. Leiva. Impact of Design Decisions in Scanpath Modeling. Proceedings of the ACM on Human-Computer Interaction, Vol. 8, issue ETRA, Article No. 228, pp 1–16.