Crossmark

A Hybrid YOLOv5-SqueezeNet Framework with Crop-Union Context Verification for Real-Time Motorcycle Rider Identification

Authors

  • Teguh Bagaskara Universitas Harapan Medan
  • Rachmat Aulia Universitas Harapan Medan

DOI:

https://doi.org/10.69916/comtechno.v4i1.410

Keywords:

Hybrid Deep Learning, YOLOv5, SqueezeNet, Crop-Union Context Verification, Edge Computing

Abstract

Motorcycles are the dominant transportation mode in developing nations, yet their high volume exacerbates traffic violations and accidents. Automated identification of motorcycle operators is crucial for Intelligent Transportation Systems (ITS). However, existing advanced vision models, such as the Segment Anything Model (SAM) or YOLOv9, demand substantial computational resources, rendering them impractical for real-time edge deployment. Furthermore, standard object detectors often fail to contextually distinguish active riders from bystanders. To address these limitations, this study proposes a lightweight, context-aware hybrid framework integrating YOLOv5 for rapid spatial detection and SqueezeNet for semantic contextual verification. The core novelty lies in the Crop-Union Context Verification Mechanism, which extracts the spatial union of paired person-motorcycle bounding boxes, expands it proportionally, and classifies the cropped region using SqueezeNet to confirm the presence of an active operator. Additionally, Contrast Limited Adaptive Histogram Equalization (CLAHE) is applied during preprocessing to enhance image quality under varying illumination. Experimental results demonstrate that the hybrid system successfully identifies motorcycle operators with high spatial confidence (YOLOv5: 0.84 for person, 0.86 for motorcycle) and robust contextual validation (SqueezeNet Top-1 prediction: moped, probability 0.247). By utilizing SqueezeNet’s highly efficient Fire Module architecture (~1.2 million parameters), the proposed framework achieves exceptional computational efficiency without sacrificing discriminative power. Ultimately, the disjunctive fusion of spatial and semantic signals ensures system resilience. This lightweight pipeline proves highly viable for real-time, AI-driven traffic surveillance on resource-constrained edge devices, offering a scalable solution for smart city infrastructure and automated law enforcement.

References

R. Artanovelia, Z. Sengaji, E. Radiansyah, and E. Ginting, “Analisis pengaruh kebiasaan, gaya hidup dan pendapatan terhadap keputusan pembelian motor Honda Vario di Kalianda,” Kalianda Halok Gagas, vol. 7, no. 1, pp. 45–55, 2024, doi: https://doi.org/10.52655/khg.v7i1.91

Y. Chang, M. Hsu, and W. Liang, “Image Sensing for Motorcycle Active Safety Warning System : Using YOLO and Heuristic Weighting Mechanism,” Sensors, vol. 25, no. 7124, pp. 1–12, 2025, doi: https://doi.org/10.52655/khg.v7i1.91

G. Kah, O. Michael, and S. Das, “Computer-vision based automatic rider helmet violation detection and vehicle identification in Indian smart city scenarios using NVIDIA TAO toolkit and YOLOv8,” Front. artficial Intell., vol. 8, no. 1582257, 2025, doi: 10.3389/frai.2025.1582257.

S. Thirumavalavan, V. Kaliyaperumal, A. Ramaiyan, and D. Pattusamy, “YOLO v9-S Net: YOLO V9 Squeeze SegNet for Object Detection Using Vehicle Image,” J. F. Robot., vol. 43, no. 3, pp. 1608–1625, 2026, doi: https://doi.org/10.1002/rob.70107.

R. Cabral, R. Santos, and J. A. F. O. Correia, “A Hybrid YOLO and Segment Anything Model Pipeline for Multi-Damage Segmentation in UAV Inspection Imagery,” Sensors, pp. 1–22, 2025, doi: https://doi.org/10.3390/s25216568

M. A. Almodhwahi and B. Wang, “A Facial-Expression-Aware Edge AI System for Driver Safety Monitoring,” Sensors, vol. 25, no. 6670, 2025, doi:https://doi.org/10.30871/jaic.v10i2.12196.

A. Subki, M. Zulpahmi, and B. Imran, “A Hybrid Framework Based on YOLOv8 and Vision Transformer for Multi-Class Detection and Classification of Coffee Fruit Maturity Levels,” vol. 9, no. 5, pp. 2019–2028, 2025, doi: 10.30871/jaic.v9i5.10590.

G. D. Deepak and S. K. Bhat, “Deep learning-based CNN for multiclassification of ocular diseases using transfer learning,” Comput. Methods Biomech. Biomed. Eng. Imaging Vis., vol. 12, no. 1, 2024, doi: 10.1080/21681163.2024.2335959.

M. S. Islami and E. R. Jamzuri, “SqueezeNet Image Embedding and Support Vector Machine for Recognizing Hand Gestures in Indonesian Sign Language System,” Ilk. J. Ilm., vol. 17, no. 2, pp. 98–106, 2025, doi: https://doi.org/10.33096/ilkom.v17i2.2476.98-106

T. Rahman et al., “Exploring the effect of image enhancement techniques on COVID-19 detection using chest X-ray images,” Comput. Biol. Med., vol. 132, no. March, p. 104319, 2021, doi: 10.1016/j.compbiomed.2021.104319.

T. B. A. Gader, H. Lachouak, S. Touil, A. K. Echi, and S. Mbarek, “Hybrid CNN-Swin Transformer for Glaucoma Screening in Fundus Images,” Proc. IEEE/ACS Int. Conf. Comput. Syst. Appl. AICCSA, no. October, 2024, doi: 10.1109/AICCSA63423.2024.10912542.

T. Y. Lin et al., “Microsoft COCO: Common objects in context,” Lect. Notes Comput. Sci. (including Subser. Lect. Notes Artif. Intell. Lect. Notes Bioinformatics), vol. 8693 LNCS, no. PART 5, pp. 740–755, 2014, doi: 10.1007/978-3-319-10602-1_48.

Y. Zhang, Y. Song, L. Zheng, O. Postolache, C. Mi, and Y. Shen, “Improved YOLOv5 Network for High-Precision Three-Dimensional Positioning and Attitude Measurement of Container Spreaders in Automated Quayside Cranes,” Sensors, vol. 24, no. 5476, 2024, doi: https:// doi.org/10.3390/s24175476.

J. Terven, D. M. Córdova-Esparza, and J. A. Romero-González, “A Comprehensive Review of YOLO Architectures in Computer Vision: From YOLOv1 to YOLOv8 and YOLO-NAS,” Mach. Learn. Knowl. Extr., vol. 5, no. 4, pp. 1680–1716, 2023, doi: 10.3390/make5040083.

Y. Zhang, Z. Guo, J. Wu, Y. Tian, H. Tang, and X. Guo, “Real-Time Vehicle Detection Based on Improved YOLO v5,” Sustain., vol. 14, no. 19, 2022, doi: 10.3390/su141912274.

X. Peng, K. Wang, Z. Zhang, N. Geng, and Z. Zhang, “A Point-Cloud Segmentation Network Based on SqueezeNet and Time Series for Plants,” J. Imaging, vol. 9, no. 258, 2023, doi: https:// doi.org/10.3390/s24175476.

R. F. Rochim, I. G. N. L. Wijayakusuma, and I. P. W. Gautama, “Performance Comparison of SqueezeNet Implementing a Dendritic Neural Model for Brain Disease Image Classification,” J. Appl. Informatics Comput., vol. 10, no. 2, pp. 1151–1158, 2026, doi: https://doi.org/10.30871/jaic.v10i2.12196.

J. Liang, C. Yang, M. Zeng, and X. Wang, “TransConver: Transformer and convolution parallel network for developing automatic brain tumor segmentation in MRI images,” Quant. Imaging Med. Surg., vol. 12, no. 4, pp. 2397–2415, 2022, doi: 10.21037/qims-21-919.

T. Mahmood, T. Saba, F. S. Alamri, A. Tahir, and N. Ayesha, “MVLA-Net: A Multi-View Lesion Attention Network for Advanced Diagnosis and Grading of Diabetic Retinopathy,” Comput. Mater. Contin., vol. 83, no. 1, pp. 1173–1193, 2025, doi: 10.32604/cmc.2025.061150.

N. H. Tasnim, S. Afrin, B. Biswas, A. A. Anye, and R. Khan, “Automatic classification of textile visual pollutants using deep learning networks,” Alexandria Eng. J., vol. 62, pp. 391–402, 2023, doi: 10.1016/j.aej.2022.07.039.

B. Imran, E. Wahyudi, A. Subki, S. Salman, and A. Yani, “Classification of stroke patients using data mining with adaboost, decision tree and random forest models,” Ilk. J. Ilm., vol. 14, no. 3, pp. 218–228, 2022, doi: 10.33096/ilkom.v14i3.1328.218-228.

Downloads

Published

2026-07-20

PlumX Metrics

Scite Metrics

Altmetric

How to Cite

Teguh Bagaskara, & Rachmat Aulia. (2026). A Hybrid YOLOv5-SqueezeNet Framework with Crop-Union Context Verification for Real-Time Motorcycle Rider Identification. Journal Computer and Technology, 4(1), 01–13. https://doi.org/10.69916/comtechno.v4i1.410

Issue

Vol. 4 No. 1 (2026): July 2026

Section

Articles

License

Copyright (c) 2026 Teguh Bagaskara, Rachmat Aulia

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.