Enhancing Video Call Safety Without Alerting Recipients

This project focuses on developing a real-time suspicious activity detection system using a pre-trained TensorFlow object detection model. The system integrates video analysis, audio processing, and email notification capabilities. It utilizes OpenCV for video stream processing and PyAudio for audio capture. A phone call simulation feature is embedded to trigger automated responses, simulating real-world security scenarios. The system is equipped with an email alert mechanism, sending notifications when certain predefined conditions are met. A graphical user interface (GUI) built with Tkinter enhances user interaction, providing an accessible and efficient tool for monitoring suspicious activities. This project showcases an advanced fusio of machine learning, computer vision, and real-time alert systems for enhanced security and monitoring applications. By leveraging real-time data processing and a simple GUI interface, this tool aids in proactive security measures and monitoring. This abstract captures the main functionalities, ensuring originality and clear articulation of the project’s objectives.