Towards Trustworthy Defense AI: Real-Time Military Asset Detection with On-Demand Explainable YOLOv8

Omaia Al-Omari; Awad Alyousef; Suliman Mohamed Fati; Esam Othman; Muhammad Rashid Naeem

doi:10.48084/etasr.17259

Authors

Omaia Al-Omari Information Systems Department, College of Computer and Information Sciences, Prince Sultan University, Riyadh, Saudi Arabia https://orcid.org/0000-0002-1638-1771
Awad Alyousef Information Systems Department, College of Computer and Information Sciences, Prince Sultan University, Riyadh, Saudi Arabia
Suliman Mohamed Fati College of Computer and Information Sciences, Prince Sultan University, Riyadh, Saudi Arabia
Esam Othman Information Systems Department, College of Computer and Information Sciences, Prince Sultan University, Riyadh, Saudi Arabia
Muhammad Rashid Naeem Department of Software Engineering, Prince Sultan University, Riyadh, Saudi Arabia

Volume: 16 | Issue: 2 | Pages: 33971-33978 | April 2026 | https://doi.org/10.48084/etasr.17259

Received: 30 December 2025 | Revised: 22 January 2026, 8 February 2026, 18 February 2026, and 22 February 2026 | Accepted: 25 February 2026 | Online: 27 February 2026
Corresponding author: Awad Alyousef

Abstract

This paper introduces Digital Shield, an interpretable deep learning framework designed for real-time military asset detection in Saudi aerial imagery. Digital Shield, a typical answer to interpretable DL models for the detection of military assets—merging YOLOv8l and SHAP to maximize high-level and transparent support for decision-making. Digital Shield is trained on the Military Assets Dataset (12 classes) after an augmentation scheme to address environmental disturbances such as desert glare. Experimental results indicate that Digital Shield outperforms both baseline models (YOLOv5 and Faster R-CNN), boasting mAP@0.5 of 92.3% and a real-time inference speed at full blast, equaling 35 fps. The proposed framework operates in two complementary modes: a continuous real-time detection pipeline and an on-demand (triggered) explainability module that is activated only during high-risk detections or upon operator request. With on-demand visual explanations, the proposed framework enhances transparency and supports operator verification, contributing to the accountable deployment of automated defense detection systems.

Keywords:

SHAP, YOLOv8, Explainable AI (XAI), UAV surveillance, real-time object detection, military asset detection

Downloads

Download data is not yet available.

References

"Saudi Arabia - Defense & Security," International Trade Administration, Jan. 03, 2024. https://www.trade.gov/country-commercial-guides/saudi-arabia-defense-security.

G. S. Xia et al., "DOTA: A Large-Scale Dataset for Object Detection in Aerial Images," in 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, June 2018, pp. 3974–3983.

A. Krizhevsky, I. Sutskever, and G. E. Hinton, "ImageNet Classification with Deep Convolutional Neural Networks," in Advances in Neural Information Processing Systems, 2012, vol. 25.

J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, "You Only Look Once: Unified, Real-Time Object Detection," in 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 2016, pp. 779–788.

M. Sohan, T. Sai Ram, and Ch. V. Rami Reddy, "A Review on YOLOv8 and Its Advancements," in Data Intelligence and Cognitive Informatics, 2024, pp. 529–545.

I. M. Shaikh, M. N. Akhtar, A. Aabid, and O. S. Ahmed, "Enhancing sustainability in the production of palm oil: creative monitoring methods using YOLOv7 and YOLOv8 for effective plantation management," Biotechnology Reports, vol. 44, Dec. 2024, Art. no. e00853.

D. Pahuja, S. Jain, and S. Kumar, "Evidence-based Inference and Quantification of Urban Expansion Using YOLOv8 and High-Resolution Satellite Imagery," Engineering, Technology & Applied Science Research, vol. 15, no. 5, pp. 26626–26631, Oct. 2025.

M. T. Ribeiro, S. Singh, and C. Guestrin, "‘Why Should I Trust You?’: Explaining the Predictions of Any Classifier," in Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, May 2016, pp. 1135–1144.

A. Jobin, M. Ienca, and E. Vayena, "The global landscape of AI ethics guidelines," Nature Machine Intelligence, vol. 1, no. 9, pp. 389–399, Sept. 2019.

S. M. Lundberg and S. I. Lee, "A Unified Approach to Interpreting Model Predictions," in Advances in Neural Information Processing Systems, 2017, vol. 30.

M. Liu and X. Di, "Extraordinary MHNet: Military high-level camouflage object detection network and dataset," Neurocomputing, vol. 549, Sept. 2023, Art. no. 126466.

M. Almutiry, "UAV Tomographic Synthetic Aperture Radar for Landmine Detection," Engineering, Technology & Applied Science Research, vol. 10, no. 4, pp. 5933–5939, Aug. 2020.

A. El-alami, Y. Nadir, and K. Mansouri, "An Efficient Geometric Transformation-Based Approach for Multi-UAV Image Stitching," Engineering, Technology & Applied Science Research, vol. 15, no. 4, pp. 25507–25513, Aug. 2025.

N. U. A. Tahir, Z. Zhang, M. Asim, S. Iftikhar, and A. A. Abd El-Latif, "PVDM-YOLOv8l: a solution for reliable pedestrian and vehicle detection in autonomous vehicles under adverse weather conditions," Multimedia Tools and Applications, vol. 84, no. 23, pp. 27045–27070, Sept. 2024.

"Military Assets Dataset (12 Classes -Yolo8 Format)." Kaggle, [Online]. Available: https://www.kaggle.com/datasets/rawsi18/military-assets-dataset-12-classes-yolo8-format.

C. Shorten and T. M. Khoshgoftaar, "A survey on Image Data Augmentation for Deep Learning," Journal of Big Data, vol. 6, no. 1, July 2019, Art. no. 60.

K. Simonyan, A. Vedaldi, and A. Zisserman, "Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps." arXiv, Apr. 19, 2014.

M. Hammad, M. ElAffendi, and S. Meshoul, "HybridFusionNet with Explanability: A Novel Explainable Deep Learning-Based Hybrid Framework for Enhanced Skin Lesion Classification Using Dermoscopic Images," Computer Modeling in Engineering & Sciences, vol. 145, no. 1, pp. 1055–1086, 2025.

S. Ren, K. He, R. Girshick, and J. Sun, "Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks," in Advances in Neural Information Processing Systems, 2015, vol. 28, Accessed: Feb. 26, 2026.

G. Jocher, "YOLOv5 by Ultralytics." Feb. 2020.

G. Tang, J. Ni, Y. Zhao, Y. Gu, and W. Cao, "A Survey of Object Detection for UAVs Based on Deep Learning," Remote Sensing, vol. 16, no. 1, Dec. 2023.