LLMER: Crafting Interactive Extended Reality Worlds with JSON Data Generated by Large Language Models

Watch a conference talk from IEEE that introduces LLMER, a framework leveraging JSON data generated by large language models to create interactive Extended Reality environments. Learn how this innovative system enables natural language interaction with virtual agents to generate objects and animations in XR. The presentation covers LLMER's multi-stage strategy for delivering contextually relevant information and its integrated modules for various XR tasks. Discover how user studies demonstrated LLMER's efficiency, showing an 80% reduction in token consumption and 60% decrease in task completion time compared to existing methods. This 11-minute session is part of the "Spatial Experiences and AI-Driven Content Generation" track at IEEE VR 2025.