Improved YOLO11-Based Weapon Detection Using Hybrid Ghost Bottleneck and Attention Module for Security Surveillance System
Adri Priadana, Duy-Linh Nguyen, Xuan-Thuy Vo, G. F. Cao, Jehwan Choi, Kang-Hyun Jo · 2025
Weapon detection plays a crucial role in enhancing intelligent security surveillance systems and ensuring safety in modern environments. A YOLO-based weapon detection model offers an effective solution for devices with limited resources, enabling real-time operation at high speeds. This work proposes a Hybrid Ghost Bottleneck (HGB) and utilizes Efficient Channel Attention (ECA) modules to improve the performance of the YOLO11-Nano version (YOLO11n). HGB allows the feature extractor to generate rich feature representations efficiently, while ECA enhances focus on essential areas. The improved YOLO11n outperforms other methods on two public weapon detection datasets. It operates at 20.07 frames per second on an Intel Core i7-9750H CPU, achieving a higher mean Average Precision and delivering faster speed than the baseline YOLO11n. This efficiency makes it well-suited for real-time security surveillance applications.