Suspicious Activity Detection

Video anomaly detection plays a critical role in surveillance, traffic monitoring, and public safety by identifying unusual events such as accidents, theft, violence, or abnormal human activities. Traditional approaches often rely on handcrafted features and statistical models, which struggle to adapt to real-time and complex scenarios. With the advancement of deep learning, object detection algorithms like YOLO (You Only Look Once) have shown remarkable efficiency in detecting and classifying objects with high accuracy and speed. This project implements a video anomaly detection framework using YOLO to process video streams and identify unusual events in real time. The system takes video input, extracts frames, processes them through YOLO for object detection, and classifies anomalies based on deviations from normal activity patterns. The results highlight the potential of YOLO based models in achieving robust, accurate, and real time anomaly detection, thereby enhancing safety and monitoring systems.