Deep Learning-Based Facial Age Estimation as a Foundation for Parental Control Systems

Abstract

This thesis explores the development of an AI-based system to estimate a child’s age from facial images, as a foundation for adaptive parental control solutions on smart devices. Motivated by the growing use of digital platforms among children and the associated risks of inappropriate content, the study aims to implement the first essential step: facial age estimation using deep learning. A convolutional neural network is trained and evaluated using a labeled dataset, and its performance is analyzed based on accuracy and mean absolute error. The findings demonstrate that deep learning can provide reliable age estimation, which may support future systems in classifying and regulating content based on user age. The results serve as a stepping stone toward intelligent, age-aware safety tools for children in digital environments.