jye9/Multi-Camera-Multi-Vehicle-Tracking-System

Multi-Camera Multi-Vehicle Tracking System (YOLO Weights)

This repository hosts the fine-tuned YOLO nano weights (e.g., 3UAVs.pt) accompanying the paper [A Topology-Aware Spatiotemporal Handover Framework for Continuous Multi-UAV Tracking](https://arxiv.org/abs/2605.15779).

These lightweight models are optimized for real-time vehicle detection from Unmanned Aerial Vehicles (UAVs) and serve as the core detection module for evaluating real-time multi-camera multi-vehicle tracking (MCMT) systems.

• —Project Page: https://www.jye.me/ICUAS2026/
• —GitHub Repository: https://github.com/JYe9/multi-camera-multi-vehicle-tracking-system
• —Dataset: jye9/Multi-Camera-Multi-Vehicle-Tracking-System

<p align="center"> <img src="https://huggingface.co/datasets/jye9/Multi-Camera-Multi-Vehicle-Tracking-System/resolve/main/assets/Pipeline.png" alt="Multi-Camera Multi-Vehicle Tracking Pipeline" width="80%"/> </p>

Citation

If you find these models, our dataset, or our framework useful in your research, please cite our paper:

@misc{ye2026topologyaware,
      title={A Topology-Aware Spatiotemporal Handover Framework for Continuous Multi-UAV Tracking}, 
      author={Jianlin Ye and Christos Kyrkou and Panayiotis Kolios},
      year={2026},
      eprint={2605.15779},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}