Paper Title
Early Threat Detection for Women’s Safety using Neural Networks and CCTV

Abstract
In public areas such as malls, banks, and parks, CCTV cameras are essential for monitoring activities and ensuring public safety. However, manually reviewing extensive footage is impractical and often leads to delayed responses to suspicious or harmful activities, particularly affecting women's safety. This project proposes an advanced automated surveillance solution utilizing the YOLOv8 neural network to enhance real-time detection of potential threats in public spaces. Building upon previous models like Convolutional Neural Networks (CNNs), which demonstrated the ability to identify risks but were limited by real-time processing and precision, YOLOv8 offers significant improvements in speed and accuracy. By automatically analysing CCTV footage to detect signs of violence or suspicious behaviour, the system generates immediate alerts, enabling prompt interventions and substantially increasing the effectiveness of security measures. This advancement aims to improve safety for women and the general public by providing a more reliable and efficient method of monitoring and responding to potential dangers in real-time. Keywords - CCTV Surveillance, YOLOv8,Neural Networks, Women’s Safety, Threat Detection.