Crafting Insights from Facial Feedback for Enhanced Understanding

This research introduces an innovative approach to facial emotion detection by integrating the Haar Cascade front face XML file for face detection and training a Convolutional Neural Network (CNN) model on the CK plus dataset for accurate emotion classification. Implemented using Jupyter Notebook and Python 3.7, our model captures facial expressions through video or webcam input, reading frames frame by frame. Each frame undergoes transformation to match the model’s input shape, facilitating precise emotion detection. The model predicts and records emotions based on facial angles, presenting the outcomes in a comprehensible pie chart upon program termination. The dataset, encompassing five emotion classes (disgust, happy, neutral, sad, surprise), forms the foundation for CNN model training, ensuring real-time emotion classification accuracy. This research not only offers a systematic flowchart for implementation but also contributes a valuable tool for emotion analysis with applications ranging from user experience design to mental health interventions.