Vehicle Re-Identification in Real Environments with Assisted Labeling and Low-Quality Data

• Artur Henrique do Nascimento Souza UFES / Atman Systems
• Augusto Abling UFES / Atman Systems
• Raquel F. Vassallo UFES

DOI: https://doi.org/10.5753/sbcup.2025.9435

Abstract

This work presents a vehicle re-identification system for ITS, focusing on Brazilian urban contexts. An open-source tool for labeling assisted by neural networks is proposed, facilitating the creation of datasets tailored to local conditions. Additionally, a spatio-temporal re-identification system is implemented, using physical metrics such as distance and speed between cameras to enhance efficiency and accuracy. Experiments show that combining local datasets with international benchmarks, such as VeRi-776, improves model generalization (mAP of 0.867). The results highlight the importance of regional adaptation for robust and effective systems in real-world scenarios.

References

Almeida, E., Silva, B., and Batista, J. (2023). Strength in diversity: Multi-branch representation learning for vehicle re-identification. In 2023 IEEE 26th International Conference on Intelligent Transportation Systems (ITSC), pages 4690–4696.

Amiri, A., Kaya, A., and Keceli, A. S. (2024). A comprehensive survey on deep-learning-based vehicle re-identification: Models, data sets and challenges.

Bewley, A., Ge, Z., Ott, L., Ramos, F., and Upcroft, B. (2016). Simple online and realtime tracking. In 2016 IEEE International Conference on Image Processing (ICIP).

Kim, H.-G., Na, Y., Joe, H.-W., Moon, Y.-H., and Cho, Y.-J. (2023). Vehicle re-identification with spatio-temporal information. In 2023 14th International Conference on Information and Communication Technology Convergence (ICTC).

Liao, H., Zheng, S., Shen, X., Li, M. J., and Wang, X. (2022). Semi-automatic data annotation system for multi-target multi-camera vehicle tracking. In 2022 IEEE 9th International Conference on Data Science and Advanced Analytics (DSAA).

Liu, X., Liu, W., Ma, H., and Fu, H. (2016). Large-scale vehicle re-identification in urban surveillance videos.

Liu, X., Liu, W., Zheng, J., Yan, C., and Mei, T. (2020). Beyond the parts: Learning multi-view cross-part correlation for vehicle re-identification. In Proceedings of the 28th ACM International Conference on Multimedia. Association for Computing Machinery.

Lv, K., Du, H., Hou, Y., Deng, W., Sheng, H., Jiao, J., and Zheng, L. (2019). Vehicle re-identification with location and time stamps. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops.

Sekachev, B. and al., E. (2020). opencv/cvat: v1.1.0.

Tang, Z., Naphade, M., Liu, M.-Y., Yang, X., Birchfield, S., Wang, S., Kumar, R., Anastasiu, D., and Hwang, J.-N. (2019). Cityflow: A city-scale benchmark for multi-target multi-camera vehicle tracking and re-identification. In 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 8789–8798.

Tzutalin (2015). Labelimg. Free Software: MIT License.

Wang, Z., Wang, L., Shi, Z., Zhang, M., Geng, Q., and Jiang, N. (2024). A survey on person and vehicle re-identification. IET Computer Vision, 18(8):1235–1268.

Weiser, M. (1999). The computer for the 21st century. SIGMOBILE Mob. Comput. Commun. Rev., 3(3):3–11.

Zhou, Y. and Shao, L. (2018). Viewpoint-aware attentive multi-view inference for vehicle re-identification. In IEEE Conference on Computer Vision and Pattern Recognition.