Research on Intelligent Applications of Metaverse Battlefield for Human-Computer Interaction
摘要
This study aims to utilize virtual reality technology to build a highly realistic metaverse battlefield based on Cesium for Unreal and Unreal Engine 5. By leveraging 3DGS rendering, Open-Sora 1.0 object-driven, and NLP multimodal interaction technologies, it achieves virtual reality multi-agent tactical confrontation decision-making deduction. The MADDPG algorithm demonstrates excellent performance in multi-agent confrontation tasks, with a win rate consistently above 95%. This research provides an efficient human-computer interaction interface (HCI) and immersive experience, holding significant application value for enhancing the tactical decision-making capabilities of military personnel.