Gamer Behavior Modeling Using Computer Vision and Artificial Intelligence

This systematic literature review examined game-player behavior modeling that integrates computer vision and artificial intelligence. A total of 50 scientific articles from the Scopus database, published between 2017 and 2025, were analyzed using the PRISMA 2020 protocol to address five research questions. The results showed that eye-tracking metrics reliably indicated cognitive load and player attention in real time. Deep learning models could predict individual-specific behavioral styles in strategic games, and data-driven analysis effectively mapped players' spatial movement dynamics in large-scale game environments. Furthermore, physiological metrics in Virtual Reality proved effective as inputs to a dynamic difficulty adjustment system, and data-driven gaze-direction modeling significantly enhanced the realism of player interactions with virtual agents. These findings confirmed that the synergy between computer vision and artificial intelligence is a crucial foundation for creating adaptive and immersive gaming experiences.