Sentiment Sounds: Orchestrating Emotions with Machine Learning for Personalized Song Recommendations

Abstract

Music has a major impact on emotions and mood management, and it frequently helps people cope with the obstacles they face on a daily basis. Finding music that perfectly captures one's present mood might be difficult, though. Currently available music recommendation systems generally rely on user preferences and listening history, which may not be useful for users looking for music to match their current feelings. This study suggests a brand-new method of providing personalised song recommendations by analysing facial expressions to ascertain the user's current mood. The system seeks to give a more individualised and organic way to finding music that corresponds with the user's emotions by utilising computer vision and machine learning techniques. Furthermore, as music may be a useful tool for self-soothing and emotional control, this technique may be especially helpful for people who are struggling with mental health concerns. In order to provide a reliable and robust music listening experience, the efficacy and accuracy of the proposed system will be thoroughly tested and user feedback on a large dataset of facial expressions and song suggestions will be collected.