This is the authors' abstract. We don't add key points for this paper.
Eye tracking data are extensively used in human-computer interaction and cognitive behavior assessments, ne-cessitating a substantial amount of labeled data for training. However, collecting substantial eye tracking data is challenging due to its variability and context dependency. In addition, a notable individual difference in the eye tracking data affects the generalization of the model. Thus, transfer learning is applied in this field to transfer the knowledge learned in one domain to a different but related domain. Transfer learning adjusts models with small-scale task data and maintains learning ability with individual differences. To the best knowledge of the authors, there has been no literature review study focusing on transfer learning applications in eye tracking research. To solve this research gap, this study describes the main transfer learning methods and explores their practical applications in eye tracking data analysis. We reviewed 29 articles from 2018 to 2024. Our research indicates that gaze classification and estimation are critical challenges in eye tracking, often addressed with transfer learning due to limited data and generalization issues. Finally, we propose technical solutions with appropriate preprocessing techniques that should be incorporated into model training and fine-tuning frameworks to enhance performance and adaptability.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Solaeman et al. (2025) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: