Enhancing Real-Time Military Object Detection in Aerial Surveillance Using a Hybrid YOLO-Transformer Architecture

DOI

10.2991/978-94-6463-940-7_23How to use a DOI?

Keywords

Real-Time Object Detection; YOLOv8; Vision Transformer; Aerial Surveillance; Military Drones; Small Object Detection; UAV

Abstract

The use of artificial intelligence in military surveillance has become essential in the rapidly changing field of modern warfare in order to improve real-time object detection in challenging aerial situations. The hybrid architecture presented in this study combines the global contextual awareness of Vision Transformers with the high-speed processing power of YOLOv8. Even while YOLOv8 provides quick and precise detection, it frequently struggles to comprehend dimensional relationships in cluttered situations, particularly when faced with small objects, occlusion, or different heights. On the contrary, vision transformers are excellent at detecting long-range dependencies and attention-based characteristics, which makes them appropriate for locating hidden or camouflaged objects. The suggested hybrid framework overcomes a number of drawbacks of traditional CNN-only or transformer-only models by combining the advantages of both models to improve detection accuracy, resilience, and inference time. The model exhibits greater performance, with higher mean Average Precision and fewer false positives, when trained and tested on aerial datasets such as DOTA and VisDrone. The significance of integrating deep learning techniques to improve situational awareness and response capabilities in defence applications is highlighted by this work. By using the proposed approach, the study advances intelligent surveillance systems that can function dependably in fast-paced, high-stakes situations when precision and speed are critical.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Drishtanta Raj Borah
AU  - J. Anju Anil
AU  - L. Andrew Wilson
PY  - 2025
DA  - 2025/12/31
TI  - Enhancing Real-Time Military Object Detection in Aerial Surveillance Using a Hybrid YOLO-Transformer Architecture
BT  - Proceedings of the Conference on Social and Sustainable Innovation in Technology & Engineering (SASI-ITE 2025)
PB  - Atlantis Press
SP  - 313
EP  - 327
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6463-940-7_23
DO  - 10.2991/978-94-6463-940-7_23
ID  - Borah2025
ER  -