SAFE VISION: Age-Based Social and Web Content -Management System to Enhance Child Safety

1Gopinath k, 2Divyesh R, 3Elangkumaran R, 4Dr T Kumanan, 5Dr K K Rehkha, 6Dr G Victo Sudha George
1,2,3 Students, 4,6 Professors-Department of CSE , 5Associate Professor- Department of CSE Dr. M.G.R
Educational and Research Institute, Maduravoyal, Chennai-95, Tamilnadu, India.

Abstract— Safe Vision is an intelligent, real-time content filtering system designed to ensure age appropriate media consumption for children andadolescents. The system dynamically analyzes and restricts unsuitable  digital content— such as images, videos, and text—across multiple platforms by leveraging advanced Computer Vision, Natural Language Processing (NLP), and Deep Learning techniques. An integrated facial recognition module, powered by tools like OpenCV and FaceNet, detects the user’s face and uses a regression model to estimate their age. Based on the estimated age, the system automatically applies customized content filtering rules. To classify and moderate sensitive content such as violence,nudity, strong language, and substance use, the system uses Convolutional Neural Networks (CNNs) implemented in TensorFlow and PyTorch. For textual analysis, transformer-based models like BERT are used to identify abusive or inappropriate language in real-time. Multimedia content is processed with tools such as FFmpeg (for video/audio handling) and OpenCV (for image/frame analysis). All sensitive data—including facial features and age—is securely handled using AES encryption via Python’scryptography library, ensuring strict privacy compliance. Safe Vision is optimized for scalable deployment in homes, schools, and publicenvironments such as digital signage systems. Its flexible design supports desktop, mobile, and edge devices, providing a robust and ethical solution for digital content safety and child well-being.Keywords — Content Filtering, Age Estimation, Facial Recognition,Computer Vision, NLP, Real-Time Moderation.