Real-Time Weapons Detection For Surviellance

This project proposes an AI-powered real-time weapon detection and alert system designed to enhance public safety and security. Using surveillance video from standard cameras, the system employs deep learning models such as YOLOv8 for weapon detection, FaceNet/DeepFace for facial recognition, and Transformer-based models for summarization of events. Whenever a weapon is detected, the system logs the incident, identifies individuals involved, and generates real-time alerts with snapshots to security staff, along with concise reports of activities. This solution reduces human monitoring fatigue, enables rapid response, and can be deployed in environments such as schools, transport hubs, malls, and defense infrastructure. By combining computer vision and NLP in one unified pipeline, the system delivers a smart, automated, and scalable approach to crime prevention and security management.