EMOTION-BASED MUSIC RECOMMENDATION SYSTEM

This paper presents an innovative Emotion-based Music Recommendation System that integrates facial emotionrecognition with personalized music selection to enhance user experience. The system employs a deep learningmodel trained to classify emotions from real-time webcam footage, detecting moods such as happiness, sadness,anger, or neutrality. Based on the detected emotion, the system dynamically selects and plays a correspondingSpotify playlist. The application leverages the Spotify API for seamless music playback, requiring a Premiumaccount for full functionality. Additionally, the system provides emotion-specific quotes and jokes to engageusers, creating a more interactive experience. Built with Streamlit for the frontend, OpenCV andKeras/TensorFlow for facial recognition and emotion classification, and Spotify's OAuth 2.0 for secureauthentication, the project highlights the potential of affective computing in entertainment and mental well-being.By processing facial data locally without storage, the system ensures user privacy while offering a unique, AIdriven music discovery tool. Future enhancements may include multi-user detection and personalized playlistcuration based on listening history.