Ai Models for Predicting Visual Attention in Digital Applications: a Comparative Pilot Analysis With Eye Tracking Results
Year: 2024
Type: conference-paper
Abstract: The paper presents a comparative pilot analysis of an AI-based visual attention prediction system with the traditional eye-tracking method in the context of UI/UX design. The study assesses the accuracy and reliability of AI-generated attention heatmaps compared to empirical eye-tracking data. The results indicate that AI systems show a strong ability to approximate human attention, achieving ROC-AUC values between 0.75 and 0.84. The study highlights AI's potential in early-stage design iterations, offering a suitable alternative to time-consuming eye-tracking testing sessions, but recommends combining AI tools with traditional methods for critical design decisions.
Authors: Neda Milić Keresteš, Gala Golubović, Sandra Dedijer, Živko Pavlović, Tamara Janjić
Institution: University of Novi Sad
Language: English
Subfield: Human-Computer Interaction
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4403644076