CampusGuardAI: An AI-Powered Real-Time Anomaly Detection System For Enhanced Campus Security

College and school campuses deserve next generation, proactive security systems that can anticipate incidents before they become rampant. Old solutions like CCTV cameras are still largely passive, recording for post facto analysis and depending on humans to watch it, which is error prone and slow. Introducing CampusGuardAI, an autonomous intelligent real-time surveillance framework driven by AI and computer vision. Built on deep learning models, YOLO for object detection and PoseNet for pose estimation, it detects outliers such as fights, trespassing, loitering in restricted areas, and abnormal behavior , including sleeping on duty. When it’s detected, the system sends automated, multi-channel alarms via a central app, SMS and email, so security response is fast. A unified dashboard aggregates data from disparate sources, giving administrators and security personnel a real-time comprehensive operational view. Designed for privacy, CampusGuardAI employs posture- and skeleton-based analysis rather than facial recognition, making it data protection-compliant. Experimental tests confirm detection accuracy and sub-second latency, demonstrating the system’s effectiveness in improving situational awareness and creating a safer campus.