Drowsiness Detection Using Convolutional Neural Networks: A Deep Learning Approach

This project develops a real-time drowsiness detection system using Convolutional Neural Networks (CNN) based on deep learning techniques. The model is trained on a dataset of facial images categorized into open eyes, closed eyes, yawning, and not yawning, which are initially captured using a web camera through a Chrome-based interface. After training, the system operates as a real-time prototype where the web camera continuously captures live images, enabling the CNN model to predict the user’s drowsiness status instantly. The system classifies eye states and yawning behavior while displaying the prediction probabilities, providing an accurate and responsive solution for drowsiness monitoring. This real-time prototype demonstrates practical applicability for enhancing safety in scenarios such as driver alertness monitoring.