Open Access

ARTICLE

EMA-GhostConv YOLOv8 Based Enhanced Vehicle Detection in Intelligent Transportation Applications

A. S. M. Masudur Rahman¹, Muhammad Zunair Zamir¹, Syed Sajid Ullah^2,*, Salman Khan², Maria Saman¹, Naqash Bahadar¹

1 School of Information Engineering, Chang’an University, Xi’an, China
2 School of Energy and Electrical Engineering, Chang’an University, Xi’an, China

* Corresponding Author: Syed Sajid Ullah. Email:

Journal on Artificial Intelligence 2026, 8, 119-136. https://doi.org/10.32604/jai.2026.076274

Received 17 November 2025; Accepted 14 January 2026; Issue published 24 February 2026

Abstract

Vehicle detection plays a pivotal role in autonomous driving, traffic monitoring, and intelligent surveillance systems. While YOLOv8 offers strong real-time performance, its detection accuracy is often limited by insufficient feature stability and suboptimal multi-scale feature fusion in complex scenes. To address these issues, we propose an enhanced YOLOv8 framework that retains the original backbone and detection head for efficiency while introducing targeted improvements to the neck architecture. Specifically, the model incorporates an Exponential Moving Average (EMA) feature layer to stabilize learning through temporally smoothed feature representations, which reduces noise and enhances generalization, and integrates GhostConv along with C3Ghost modules to enrich feature diversity with minimal computational overhead. We conduct comprehensive experiments on both a custom vehicle dataset and the KITTI benchmark, demonstrating consistent and significant improvements over the baseline YOLOv8. On KITTI, our method achieves gains of +11.68% in precision, +12.84% in recall, +10.20% in mAP@50, and +12.73% in mAP@50–95. Furthermore, ablation studies validate the individual contribution of each component, and comparisons with multiple state-of-the-art detectors confirm the competitiveness of our approach in both accuracy and efficiency. The proposed EMA-GhostConv YOLOv8 framework thus offers a robust, lightweight, and high-performance solution for vehicle detection in intelligent transportation applications.

---

Keywords

---