EyeFormer: Predicting Personalized Scanpaths with Transformer-Guided Reinforcement Learning

要旨

From a visual-perception perspective, modern graphical user interfaces (GUIs) comprise a complex graphics-rich two-dimensional visuospatial arrangement of text, images, and interactive objects such as buttons and menus. While existing models can accurately predict regions and objects that are likely to attract attention ``on average'', no scanpath model has been capable of predicting scanpaths for an individual. To close this gap, we introduce EyeFormer, which utilizes a Transformer architecture as a policy network to guide a deep reinforcement learning algorithm that predicts gaze locations. Our model offers the unique capability of producing personalized predictions when given a few user scanpath samples. It can predict full scanpath information, including fixation positions and durations, across individuals and various stimulus types. Additionally, we demonstrate applications in GUI layout optimization driven by our model.

著者
Yue Jiang
Aalto University, Espoo, Finland
Zixin Guo
Aalto University, Espoo, Finland
Hamed Rezazadegan Tavakoli
Nokia Technologies, Espoo, Finland
Luis A.. Leiva
University of Luxembourg, Esch-sur-Alzette, Luxembourg
Antti Oulasvirta
Aalto University, Helsinki, Finland
論文URL

https://doi.org/10.1145/3654777.3676436

動画

会議: UIST 2024

ACM Symposium on User Interface Software and Technology

セッション: 3. Machine Learning for User Interfaces

Westin: Allegheny 3
5 件の発表
2024-10-15 18:00:00
2024-10-15 19:15:00