NewsEye: AI Powered Anomaly Detection and News Reporting

Traditional methods of anomaly detection frequently struggle with real-time monitoring, resulting in delayed detection of odd events. To address these limitations, we propose an integrated system that pre-processes surveillance video with resolution correction and noise reduction, followed by feature extraction using pre-trained CNN models like ResNet. Density-based methods, such as the Local Outlier Factor (LOF), are used to identify anomalies, which are then analysed using AUC-ROC criteria. GPT and other Natural Language Processing (NLP) models will also use to automate the generation of extensive news stories. The NewsEye project takes a fresh way to addressing the issues provided by the massive amounts of data produced by modern surveillance technologies. The proposed system intends to improve surveillance efficiency by delivering real-time anomaly detection and automatic news reporting, decreasing the need for human intervention.