Dasam Srinivas,
A.V.S. Pawan Kumar,
- PG Scholar, Department of CSE Gandhi Institute of Engineering and Technology University, Odisha, India
- Associate Professor, Department of CSE Gandhi Institute of Engineering and Technology University, Odisha, India
Abstract
Accurate object detection is essential for autonomous driving systems, where vehicles must continuously perceive surrounding road users and infrastructure in real time. Although the You Only Look Once (YOLO) family of deep learning detectors has achieved significant success due to its high speed and efficiency, detecting small or partially occluded objects in complex traffic environments remains challenging. This paper presents an attention-enhanced YOLO framework designed to improve feature representation and detection reliability for autonomous driving applications. The proposed method integrates spatial and channel attention mechanisms into the detection architecture, enabling the network to focus on informative regions of the scene while suppressing irrelevant features. The model is evaluated using representative autonomous driving datasets, including KITTI and BDD100K, and compared with existing YOLO variants. Experimental results demonstrate that the proposed framework achieves precision of 0.97, recall of 0.92, and mean average precision ([email protected]) of 0.982, outperforming baseline models such as YOLOv5 and YOLOv8. In addition, the system maintains real-time performance of approximately 94 frames per second, making it suitable for deployment in practical autonomous vehicle perception systems. The results indicate that incorporating attention mechanisms significantly enhances detection accuracy, particularly for small and partially occluded objects, while preserving the computational efficiency required for real-time driving environments.
Keywords: Autonomous vehicles, object detection, YOLO, attention mechanism, deep learning, real-time detection, computer vision, autonomous driving perception
[This article belongs to Journal of Multimedia Technology & Recent Advancements ]
References
- Wang CY, Bochkovskiy A, Liao HYM. YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition; 2023. p. 7464-75. doi:10.1109/CVPR52729.2023.00721.
- Ultralytics YOLO26, YOLO11, YOLOv8: object detection, instance segmentation, semantic segmentation, image classification, pose estimation, object tracking [Internet]. GitHub; 2017. Available from: https://github.com/ultralytics/ultralytics
- Wang CY, Yeh IH, Liao HYM. YOLOv9: Learning what you want to learn using programmable gradient information. In: European Conference on Computer Vision. Cham: Springer Nature Switzerland; 2024. p. 1-21.
- Tian Y, Ye Q, Doermann D. YOLOv12: Attention-centric real-time object detectors. Adv Neural Inf Process Syst. 2026;38:78433-57.
- Carion N, Massa F, Synnaeve G, Usunier N, Kirillov A, Zagoruyko S. End-to-end object detection with transformers. In: European Conference on Computer Vision. Cham: Springer International Publishing; 2020. p. 213-29.
- Tamilarasi T, Muthulakshmi P, Miraei Ashtiani SH. Improved YOLO-based real-time brinjal detection algorithm for vision modules in harvesting robots. Eng Res Express. 2025;7(3):035234. doi:10.1088/2631-8695/adef00.
- Ganguly B, Dey D, Munshi S. An attention deep learning framework-based drowsiness detection model for intelligent transportation system. IEEE Trans Intell Transp Syst. 2025;26:4517-27. doi:10.1109/TITS.2025.3544138.
- Liang E, Wei D, Li F, Lv H, Li S. Object detection model of vehicle-road cooperative autonomous driving based on improved YOLO11 algorithm. Sci Rep. 2025;15:32348. doi:10.1038/s41598-025-18263-9. PMID:40897806.
- Wei F, Wang WW. SCCA-YOLO: a spatial and channel collaborative attention enhanced YOLO network for highway autonomous driving perception system. Sci Rep. 2025;15:6459. doi:10.1038/s41598-025-90743-4. PMID:39987326.
- Dustali A, Hasanlou M, Azimi SM. Comparative analysis of YOLO-based algorithms for vehicle detection in aerial imagery. Int Arch Photogramm Remote Sens Spat Inf Sci. 2025;XLVIII-G:411-6. doi:10.5194/isprsarchives-XLVIII-G-2025-411-2025.
- Liu J, Zhan J, Zhang J, Chen J, Song Y, Tang L, et al. Robust scale fusion and edge-aware feature attention network for remote sensing UAV road detection under harsh weather. Results Eng. 2025;27:106172. doi:10.1016/j.rineng.2025.106172.
- Devgan MS, Singh G, Sharma P, Kumar T, Cheng X, Ahlawat D. Attention-guided lightweight CNN-transformer fusion for real-time traffic sign recognition in adverse environments: HACTNet. IET Intell Transp Syst. 2026;20:e70167. doi:10.1049/itr2.70167.
- Hameed R, HN M, P CS, Oggu VB, B I. Domain-adaptive reinforcement learning framework for robust autonomous vehicle navigation. In: Proceedings of the 2nd International Conference on Intelligent Algorithms for Computational Intelligence Systems (IACIS); 2025. p. 1-6. doi:10.1109/IACIS65746.2025.11210918.
- Geiger A, Lenz P, Urtasun R. Are we ready for autonomous driving? The KITTI vision benchmark suite. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition; 2012. p. 3354-61. doi:10.1109/CVPR.2012.6248074.
- Yu F, Chen H, Wang X, Xian W, Chen Y, Liu F, et al. BDD100K: a diverse driving dataset for heterogeneous multitask learning. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition; 2020. p. 2633-42. doi:10.1109/CVPR42600.2020.00271.
Journal of Multimedia Technology & Recent Advancements
| Volume | 13 | |
| Issue | 02 | |
| Received | 02/07/2026 | |
| Accepted | 11/08/2026 | |
| Published | 16/08/2026 | |
| Publication Time | 45 Days |