This repository serves as a comprehensive resource for researchers exploring the intersection of Unmanned Aerial Vehicles (UAVs) and Large Language Models (LLMs). It provides curated tables summarizing LLMs, Vision-Language Models (VLMs), Vision-Foundation Models (VFMs), UAV configurations, swarm algorithms, and extensive datasets across various application domains relevant to UAVs. The primary goal is to offer a structured overview and facilitate research in agentic low-altitude mobility.

How It Works

The repository is structured around a survey paper, "UAVs Meet LLMs: Overviews and Perspectives Toward Agentic Low-Altitude Mobility," and contains detailed tables that categorize and list relevant models, datasets, and techniques. It covers foundational aspects of UAVs, LLM/VLM/VFM advancements, and specific application areas like visual perception, navigation, planning, and flight control, providing a structured knowledge base for the field.

Quick Start & Requirements

This repository is primarily a curated collection of information and does not require installation or execution. It serves as a reference guide.

Highlighted Details

• Extensive tables detailing LLMs, VLMs, and VFMs, including major models from OpenAI, Google, Meta, and Anthropic.
• Comprehensive lists of UAV-oriented datasets for environmental perception, object tracking, action recognition, and navigation, with details on types and amounts.
• Categorization of Foundation Model-based UAV systems across visual perception, VLN, planning, and flight control tasks.
• Overviews of UAV swarm path planning, task allocation, communication architectures, and formation control algorithms.

Maintenance & Community

The repository is actively maintained by Yonglin Tian and Fei Lin, with contributions from Yiduo Li, Tengchao Zhang, and Xuan Fu. Users are encouraged to open issues or pull requests for suggestions and contributions.

Licensing & Compatibility

This project is licensed under the MIT License, permitting commercial use and integration into closed-source projects.

Limitations & Caveats

The repository is a curated list and does not provide code implementations or direct access to the listed models or datasets. Users will need to refer to the original sources for practical application.