Weapon detection with audio alert using efficientdet

In recent years, the rising incidence of violence in public spaces has underscored the urgent need for intelligent surveillance systems. This project proposes a real-time weapon detection system that leverages the EfficientDet object detection model to identify potential threats from camera feeds, including weapons such as Heavy Gun, Knife, Pistol, and sharp objects like "pisau". The system is capable of accurately detecting both individuals (Person) and the presence of weapons using a custom-trained EfficientDet model on a labeled dataset. Upon detecting a weapon, the system immediately triggers an audio alert, allowing timely human intervention. The frontend is designed using HTML and CSS, providing a user-friendly interface for real-time monitoring and alert notifications. The integration of EfficientDet ensures a balance between speed and accuracy, making the solution suitable for deployment in schools, malls, public transport hubs, and other high-risk environments. This project demonstrates the potential of combining deep learning-based object detection with real-time alert mechanisms to enhance public safety and proactive threat response. Keywords EfficientDet, Weapon Detection, Object Detection, Audio Alert System, Real-time Surveillance, Public Safety, Deep Learning, Security AI, HTML, CSS, Person Recognition, Knife, Gun, Pistol Detection.