Key points are not available for this paper at this time.
With the rapid advances in rendering technology, the implementation of highly realistic avatars in a virtual reality (VR) environment has become a possibility. Such realistic avatars have significant effects on the user's avatar perception and interactive experience. However, they consume excessive computing resources and cause the uncanny valley phenomenon, which can degrade the user experience, necessitating extensive research on the visual realism of realistic avatars. Against this backdrop, this study performs an in-depth analysis on the effect of the visual realism of avatars on the user's avatar perception in a VR environment, thereby adjusting visual realism through various factors and evaluating its effect on avatar perception. The aim is to investigate the optimal level of visual realism that allows for efficient utilization of computing resources while allowing avatar perception. Unlike previous studies that mainly focused on visual realism in two-dimensional environments, this study considers the characteristics of VR environments and investigates the various levels of visual realism that affect avatar perception in an immersive environment. In this study, the MetaHuman Creator of Unreal Engine was used to create highly realistic personalized avatars. An experiment was conducted with 40 participants recruited to evaluate the effect of visual realism on avatar perception in both VR head-mounted display (HMD) and two-dimensional display environments. This study contributes significantly to the field of personalized avatar research by providing more comprehensive analysis and insights. The analysis results indicate that the immersive environment has a significant effect on avatar perception. In particular, high avatar perception was observed in the VR HMD environment, implying that enhanced immersive experiences are important in virtual reality. The study also examined the impact of texture and the level of detail on avatar perception. The results show that there is no significant difference between moderate and high levels of detail, which indicates that when designing avatars in virtual reality, greater visual realism may not always lead to higher avatar perception. The findings of this study suggest a new direction for efficient utilization of computing resources and enhanced user experiences. It is further expected that these findings will present a standard for rendering avatars in virtual reality applications.
Lee et al. (Tue,) studied this question.