Reproduce “Navigation Pixie: Implementation and Empirical Study Toward On-demand Navigation Agents in Commercial Metaverse” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Hikari Yanagawa, Yuichi Hiroi, Satomi Tokida, Yuji Hatada, Takefumi Hiraki
The paper introduces "Navigation Pixie," an LLM-powered on-demand navigation agent for commercial metaverse platforms, designed with a loosely coupled architecture for broad compatibility. Empirical studies on the Cluster platform show the agent increases user dwell time by 1.5-1.7 times and free exploration time by 3-5 times, fostering deeper engagement and providing contextual insights for users in diverse virtual environments.
Key insights
- LLM-based on-demand agents significantly enhance user engagement and autonomy by enabling flexible conversations and adapting to evolving user interests in complex virtual environments.
- A loosely coupled architecture effectively bridges the gap between academic AI advancements and commercial metaverse platforms, enabling scalable deployment and evaluation.
- The effectiveness and social perception of intelligent agents are influenced by the access platform (PC vs. VR-HMD) and the environmental context (e.g., social vs. exhibition spaces).
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.