A Web-Based Facial Recognition Application Using Convolutional Neural Networks (CNNs)

Facial recognition has become one of the most fascinating areas of applied artificial intelligence, particularly due to its ability to connect human perception with machine-level precision. This paper introduces a web application that allows users to upload personal images and receive real-time interpretations of facial characteristics by comparing facial structure to a curated set of celebrity portraits. The core of the proposed system is powered by a Convolutional Neural Network (CNN) trained specifically for facial analysis. Once an image is submitted, it is processed, validated and passed to the model, which performs feature extraction and returns a meaningful output in seconds. The model itself is optimized using techniques derived from modern CNN architectures, with a focus on balancing performance. From a technical standpoint, the frontend of the application is built using React and TypeScript. On the backend, a Python-based server handles image processing and communicates with the AI model.