Integrating Facial Expression Recognition for Emotion Responsive Music Recommendations

Music has a profound impact on human emotions, enriching experiences and serving as a powerful means of expression. In contemporary times, the synergy between music and machine learning has revolutionized the way we engage with and explore musical experiences. This research delves into the intricate relationship between emotions and music, leveraging machine learning algorithms, culminating in an innovative emotion-based music recommendation system. This system employs facial emotion recognition (FER) as a key mechanism, harnessing the expressive canvas of the human face to decipher emotions accurately. The system integrates YOLOv4 for facial detection and the Humingbird library for emotion analysis. Leveraging a dataset comprising 278,000 Spotify songs labeled by emotion, the system recommends a Spotify song URL aligned with the user’s detected emotion. We also created a simple yet efficient interface for user interaction utilising streamlit web application tools. This system not only underscores the intricacies of emotional connections with music but also underscores the transformative potential of combining human expression with cutting-edge technology.