MC-GTA: A Synthetic Benchmark for Multi-Camera Vehicle Tracking

Dataset

The MC-GTA dataset is designed for multi-camera vehicle tracking (MCVT) in urban environments, crucial for city-scale traffic analysis, management, and security applications. Traditional MCVT systems face challenges due to the scarcity of annotated data necessary for training and testing deep learning-based computer vision models. To address this, the MC-GTA dataset offers a synthetic collection of urban scene images captured from the virtual environment of the Grand Theft Auto 5 (GTA) video game. This dataset features recordings from multiple cameras placed at various crossroads, with automatically generated annotations including bounding boxes and unique vehicle IDs consistent across different video sources. The dataset aims to provide a valuable benchmark for MCVT tasks, demonstrating its utility through performance evaluation with a state-of-the-art MCVT approach. Additionally, the dataset and tools for creating custom scenarios are publicly accessible at https://github.com/GaetanoV10/GT5-Vehicle-BB.

Citing the MC-GTA

The MC-GTA is released under a Creative Commons Attribution license, so please cite the MC-GTA if it is used in your work in any form.Published academic papers should use the academic paper citation for our MC-GTA paper

@inproceedings{ciampi2023mc,
title={Mc-gta: A synthetic benchmark for multi-camera vehicle tracking},
author={Ciampi, Luca and Messina, Nicola and Valenti, Gaetano Emanuele and Amato, Giuseppe and Falchi, Fabrizio and Gennaro, Claudio},
booktitle={International Conference on Image Analysis and Processing},
pages={316--327},
year={2023},
organization={Springer}
}

Personal works, such as machine learning projects/blog posts, should provide a URL to the MC-GTA Zenodo page (https://doi.org/10.5281/zenodo.5996890), though a reference to our MC-GTA paper would also be appreciated.

Contact Information

If you would like further information about the MC-GTA or if you experience any issues downloading files, please contact us at luca.ciampi[at]isti.cnr.it

Acknowledgements

Supported by: MOST - Sustainable Mobility National Research Center, funded by the European Union Next-GenerationEU (Piano Nazionale di Ripresa E Resilienza (PNRR) - Missione 4 Componente 2, Investimento 1.4 - D.D. 1033 17/06/2022, CN00000023); AI4Media – A European Excellence Centre for Media, Society, and Democracy (EC, H2020 No. 951911); SUN – Social and hUman ceNtered XR (EC, Horizon Europe No. 101092612).