Utkarsh-Mishra444/Sparsely-Grounded-Long-Range-Navigation

AgentNav: Zero-shot sparsely grounded long-range visual navigation in real-world cities using Multimodal Large Language Models (MLLMs).

35
/ 100
Emerging

This project helps urban planning researchers and autonomous navigation system designers understand how AI agents can navigate complex city environments using only visual cues. It takes street-level images from intersections and a destination, then outputs a sequence of navigation decisions and an estimated path, acting like a human driver or pedestrian navigating without GPS or maps. It's designed for those who need to simulate or evaluate advanced visual navigation capabilities in diverse real-world urban settings.

Use this if you are exploring advanced AI techniques for self-localization and pathfinding in real-world urban environments without relying on GPS, maps, or explicit instructions.

Not ideal if you need a simple, ready-to-deploy GPS-based navigation system or a tool for navigating with pre-defined maps and landmarks.

urban-robotics autonomous-navigation visual-pathfinding city-simulation AI-geospatial-reasoning
No License No Package No Dependents
Maintenance 13 / 25
Adoption 4 / 25
Maturity 5 / 25
Community 13 / 25

How are scores calculated?

Stars

8

Forks

2

Language

Python

License

Last pushed

Mar 17, 2026

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/computer-vision/Utkarsh-Mishra444/Sparsely-Grounded-Long-Range-Navigation"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.