Suspicious Human Activity Prediction Using CNN

This computer vision project applies Convolutional Neural Networks to detect suspicious human activities in visual data, addressing the growing need for automated security monitoring systems.

The system is designed to identify potentially dangerous or criminal behavior patterns in real-time, helping to reduce crime incidents and mob violence through early detection and alert mechanisms.

Incorporating explainable AI techniques using GradCAM, the system provides visual explanations for its predictions, making it transparent and trustworthy for security personnel and law enforcement.