CVPR 2026 Workshop · GRAIL-V

LLM-Guided Agentic Floor Plan Parsing for Accessible Indoor Navigation of Blind and Low-Vision People

University of Maryland, Baltimore County

Overview of the agentic pipeline showing the five-agent workflow and three-tier RAG knowledge store.
Figure 1. Overview of our agentic pipeline showing the five-agent workflow and three-tier RAG knowledge store. Original figure from the arXiv HTML version of the paper.

Abstract

Indoor navigation remains a critical accessibility challenge for the blind and low-vision (BLV) individuals, as existing solutions rely on costly per-building infrastructure. We present an agentic framework that converts a single floor plan image into a structured, retrievable knowledge base to generate safe, accessible navigation instructions with lightweight infrastructure. The system has two phases: a multi-agent module that parses the floor plan into a spatial knowledge graph through a self-correcting pipeline with iterative retry loops and corrective feedback; and a Path Planner that generates accessible navigation instructions, with a Safety Evaluator agent assessing potential hazards along each route. We evaluate the system on the real-world UMBC Math and Psychology building (floors MP-1 and MP-3) and on the CVC-FP benchmark. On MP-1, we achieve success rates of 92.31%, 76.92%, and 61.54% for short, medium, and long routes, outperforming the strongest single-call baseline (Claude 3.7 Sonnet) at 84.62%, 69.23%, and 53.85%. On MP-3, we reach 76.92%, 61.54%, and 38.46%, compared to the best baseline at 61.54%, 46.15%, and 23.08%. These results show consistent gains over single-call LLM baselines and demonstrate that our workflow is a scalable solution for accessible indoor navigation for BLV individuals.

Framework

01

Multi-agent parsing

The Parser extracts rooms and adjacencies, the Graph Builder constructs a validated floor-plan graph, and the Self-Critic enforces structural checks with feedback loops.

02

Safety Evaluator

An agent flags hazards along each route, such as narrow passages and dead ends.

03

Three-tier RAG knowledge store

Results are stored as graph, vector, and visual tiers for retrieval.

04

Path Planner

The RAG Navigator retrieves context and the Planner generates BFS-based navigation with landmarks and safety notes, re-routing on hazard alerts.

Resources

Preprint
arXiv:2604.23970
Open access (CVF)
CVPRW 2026 paper (PDF)
DOI
10.48550/arXiv.2604.23970

BibTeX

@inproceedings{ayanzadeh2026llm,
  title = {LLM-Guided Agentic Floor Plan Parsing for Accessible Indoor Navigation of Blind and Low-Vision People},
  author = {Ayanzadeh, Aydin and Oates, Tim},
  booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops},
  year = {2026},
  eprint = {2604.23970},
  archivePrefix = {arXiv},
  primaryClass = {cs.AI},
  doi = {10.48550/arXiv.2604.23970},
  url = {https://arxiv.org/abs/2604.23970}
}