A Comparative Study of Weapon Detection in Images Using Deep Learning Algorithms

Finding weapons is an essential part of maintaining public safety and averting violent happenings. The YOLO (You Only Look Once) object detection algorithm has become a well-liked and effective method for real-time firearm identification using computer vision algorithms in recent years. The application of YOLO for weapon detection and its efficacy in detecting firearms, knifes, and other weapons are covered in this report. Using a collection of annotated photos, we trained a YOLOv8 model, and then we assessed its performance using a test dataset. The findings show that YOLO is an accurate and reliable approach for weapon detection, improving the model's performance in many settings and obtaining a high mAP and mA.