Avatar-Driven Interaction: Constructing Immersive Education in Virtual Classrooms

With advances in multimedia and computer graphics technology, 3D virtual avatars have demonstrated promising performance in education. However, existing research faces challenges in systematically studying student interaction. In this study, a Virtual Classroom Interaction and Modal Assessment (VCIMA) is proposed to integrate various factors to explore the impact on learners’ sense of presence, engagement, and preference. Specifically, we analyze key digital avatar types, learning perspectives, viewing devices, and their roles in shaping immersive learning experiences. Then, five elaborate scenarios around a physical demonstration are introduced to simulate diverse learning contexts, assess multimodal interactions, and evaluate how different instructional designs influence learner responses. Subsequently, a substantial corpus of experiments has been conducted under different conditions to verify the effectiveness of VCIMA and reveal that the adaptive multimodal feedback of 3D avatars in virtual reality enhances participation and presence. These findings provide actionable insights for designing future virtual learning environments, emphasizing the necessity of personalized interaction models.