Music Player Based On Real Time Emotion Detection Using CNN

Abstract

A person’s music preferences are shaped not only by their current emotional state but also by their historical musical choices. This research presents a robust music recommendation system that proposes music aligned with the user’s present mood. Recognizing the profound impact of harmonious sounds on emotions, the system goes a step further by dynamically tailoring playlist creation to the listener’s experience, leveraging real-time emotional cues. The incorporated CNN model adeptly examines visual and audio content from radio waves, discerning emotions such as pleasure, sadness, enthusiasm, and tranquility. Users actively contribute to refining the system through feed- back, fostering an interactive and evolving emotional connection between the listener and their curated collection. This endeavor marks a significant advancement in cultivating a more empathetic and immersive playback experience, strengthening the emotional bond between users and their musical selections. The music classification engine achieves noteworthy accuracy by employing audio elements for precise categorization, collectively enhancing the user’s musical journey through the seamless integration of mood-based recommendations and accurate music classification.