Xpressify: Facial Emotion Recognition and Music Recommendation System
Devi R., Balasakthi, Harish, Maithireyan, Vendhan · Recent Research Reviews Journal · 2025
Facial Emotion Detection and Music Recommendation System is a real-time facial emotion detection and personalized music recommendation system. It employs a trained Convolutional Neural Network (CNN) trained on the FER2013 database for seven emotion recognitions, i.e., anger, disgust, fear, happiness, sorrow, surprise, and neutrality. Everything is accomplished through Artificial Intelligence (AI) and Computer Vision technology. React and Flask are used to create the frontend and backend, while OpenCV is used to detect the facial emotion of the user through the webcam. The system does not pre-downloading playlists but dynamically uses the YouTube Data Application Programming Interface (API) to download appropriate music tracks depending on the emotional state of the user. Measures of precision, recall, accuracy (79.07%), and loss values are used to deploy the proposed system. The values indicate the viability of the application of Xpressify for use in real-time affect classification and customized music. With dynamic, interactive, and emotive-response music recommendations, the integration provides a more realistic experience for the user and is applicable in entertainment, mental health, and human-computer interaction.