Emotion-Aware Music Recommendation

Music as a universal language has been profoundly ingrained with human emotions over years-capable of influencing mood, decreasing stress, and enhancing mental function. Traditional music recommendation systems are predominantly founded on static likes, history of previous listening, or collaborative filtering and sometimes fail to discern the user's true emotional requirements in real-time. To fill this gap, we propose a new system called "Emotion-Aware Music Recommendation System" that leverages realtime emotion recognition using deep learning techniques to dynamically choose music playlists to correspond to the emotional state of the user at a particular moment. With the addition of Convolutional Neural Networks (CNNs) to face emotion recognition and an emotion-to-music mapped selection algorithm, the system depicts a human-like methodology towards increased user satisfaction and emotional bliss. Experimental results validate a gross emotion classification accuracy of above 85%, and high user acceptance of the dynamic, empathetic music recommendation experience