Guide-LLM is an innovative embodied LLM agent that leverages text-based topological mapping to provide precise navigation assistance for persons with visual impairments (PVI), addressing limitations of traditional mobility aids.

• Combines large language models with topological mapping to deliver detailed spatial information
• Creates a text-based representation of environments that LLMs can process and reason with
• Enables natural language interaction for intuitive guidance through complex spaces
• Provides contextually aware directions that adapt to the user's needs

This research represents a significant advancement in assistive medical technology, offering potential for greater independence and spatial understanding for millions with visual impairments, while reducing reliance on human guides or service animals.

Guide-LLM: An Embodied LLM Agent and Text-Based Topological Map for Robotic Guidance of People with Visual Impairments