Personalized Image Preference Assessment for Individuals with Color Vision Deficiency

Individuals with Color Vision Deficiency (CVD) often experience contrast loss, which affects their visual perception in daily life. Various compensation methods have been developed to enhance image contrast while maintaining naturalness. However, evaluating these compensation images poses a significant challenge due to the subjective nature of perception and individual differences in visual deficiency. To tackle this issue, we propose a three-phase framework for personalized image preference assessment: (1) Generic aesthetics modeling,(2) CVDspecific fine-tuning, and (3) Personalized preference modeling. Objective experiments validate the effectiveness of our approach, demonstrating its ability to accurately model individual preferences for CVD individuals.