Split Computing for Vehicle Detection in Smart Parking

• Milena F. Silva UNICAMP
• Heitor H. da Silva UNICAMP
• Mateus C. Oliveira UNICAMP
• Carlos Senna UNICAMP
• Allan M. de Souza UNICAMP
• Juliana F. Borin UNICAMP
• Luiz F. Bittencourt UNICAMP

DOI: https://doi.org/10.5753/courb.2026.23194

Resumo

The increasing complexity of deep learning models like YOLO challenges the capabilities of low-cost edge devices in smart parking systems. To address this gap, we propose a Split Computing architecture applied to vehicle detection, where the neural network is partitioned between an edge client and a central server. This design offloads intensive computation, reducing local inference time and resource usage while preserving model accuracy. We evaluated distinct partitioning configurations, assessing computational costs and communication latency. Results demonstrate the solution’s effectiveness, highlighting its scalability for multi-tenant environments and sustainability potential by enabling the repurposing of low-cost TV set-top boxes as functional edge nodes.

Referências

Baggio, J. V., Gonzalez, L. F. G., and Borin, J. F. (2020). Smartparking a smart solution using deep learning. [link].

Chakroun, I., Vander Aa, T., Wuyts, R., and Verachtert, W. (2021). Distributing intelligence for object detection using edge computing. In 2021 IEEE 14th International Conference on Cloud Computing (CLOUD), pages 681–687.

da Luz, G., Sato, G., Bannwart, T., Gonzalez, L., and Borin, J. (2025a). 10 years of deep learning for vehicle detection at a smart parking : What has changed? In Anais do IX Workshop de Computação Urbana, pages 127–140, Porto Alegre, RS, Brasil. SBC.

da Luz, G. P. C. P., Sato, G. M., Gonzalez, L. F. G., and Borin, J. F. (2025b). Repurposing of tv boxes for a circular economy in smart cities applications. Scientific Reports, 15(1):22638.

Eshratifar, A., Abrishami, A., and Pedram, M. (2019). Bottlenet: A deep learning architecture for intelligent mobile cloud offloading. In 2019 IEEE/ACM International Symposium on Low Power Electronics and Design (ISLPED), pages 1–6. IEEE.

Gupta, O. and Raskar, R. (2018). Distributed learning of deep neural network over multiple agents. Journal of Network and Computer Applications, 116:1–8.

Kang, Y., Hauswald, J., Gao, C., Rovinski, A., Mudge, T. N., Mars, J., and Tang, L. (2017). Neurosurgeon: Collaborative intelligence between the cloud and mobile edge. In Proceedings of ASPLOS ’17, pages 615–629.

Marinova, M. and Rakovic, V. (2024). Accelerating convergence in split learning for time-varying and resource-limited environments. In 2024 IEEE 22nd Mediterranean Electrotechnical Conference (MELECON), pages 13–18.

Matsubara, Y. and Levorato, M. (2020). Split computing for complex object detectors: Challenges and preliminary results. In Proceedings of the 4th ACM Workshop on Hot Topics in Video Analytics and Intelligent Edges (VideoAI ’20).

Matsubara, Y. and Levorato, M. (2022). Split computing and early exiting for deep learning applications: Survey and research challenges. ACM Computing Surveys.

Nations, U. (2015). Transforming our world: the 2030 agenda for sustainable development. [link].

Neduchal, P., Straka, J., Sieber, M., and Gruber, I. (2024). Comparison of split computing scenarios for object detection. In IFAC-PapersOnLine (18th IFAC Conference on Intelligent Manufacturing Systems), pages 120–125.

Okano, M. T., Lopes, W. A. C., Ruggero, S. M., Vendrametto, O., and Fernandes, J. C. L. (2025). Edge AI for industrial visual inspection: YOLOv8-based visual conformity detection using raspberry pi. Algorithms, 18(8).

P C P da Luz, G., Massuyoshi Sato, G., Fernando Gomez Gonzalez, L., and Freitag Borin, J. (2026). Smart parking with pixel-wise ROI selection for vehicle detection using yolov8, yolov9, yolov10, and yolov11. Internet of Things, 36:101858.

Rey, L., Bernardos, A. M., Dobrzycki, A. D., Carramiñana, D., Bergesio, L., Besada, J. A., and Casar, J. R. (2025). A performance analysis of you only look once models for deployment on constrained computational edge devices in drone applications. Electronics, 14(3).

Sato, G. M., da Luz, G. L. N., Gonzalez, L. F. P., and Borin, J. (2024). Reaproveitamento de tv boxes para aplicação de contagem de pessoas na borda em cidades inteligentes. Anais do VIII Workshop de Computação Urbana (CoUrb 2024).

Shoup, D. (2021). Pricing curb parking. Transportation Research Part A: Policy and Practice, 154:399–412.

Wu, W., Li, M., Qu, K., Zhou, C., Shen, X., Zhuang, W., Li, X., and Shi, W. (2023). Split learning over wireless networks: Parallel design and resource management. IEEE Journal on Selected Areas in Communications, 41(4):1051–1066.

Yao, J. (2023). Split learning for image classification in internet of drones networks. In 2023 IEEE 24th International Conference on High Performance Switching and Routing (HPSR), pages 52–55.

Zhang, L., Chen, L., and Xu, J. (2021). Autodidactic neurosurgeon: Collaborative deep inference for mobile edge intelligence via online learning. In Proceedings of The Web Conference (WWW) 2021, pages 3111–3123.