メタバースコミュニケーションにおける生成AIを用いた可視化による概念の共有

抄録

As the metaverse expands, virtual communication presents opportunities for global interaction but also poses challenges in conveying nuanced or abstract ideas through language alone. To address this, we developed a real-time visualization system using AI technology to enhance communication within the metaverse. The system employs the multi-space design model, considering value, meaning, state, and attribute spaces to support seamless communication and effective information sharing among users with diverse backgrounds. We conducted two experiments to evaluate the system's effectiveness. The first experiment tested the system's ability to accurately visualize user-expressed concepts, showing improved accuracy through feedback, though initial results indicated a need for further algorithmic optimization. The second experiment assessed the system's role in enhancing communication quality, confirming its utility in facilitating mutual understanding and concept concretization. These findings suggest that AI-driven visualization systems could become essential tools in the metaverse, especially as global collaboration continues to grow. Future research will focus on refining system accuracy, reducing response times, and broadening user applicability.