Music recommendation based on face emotion recognition

Abstract

This paper proposed a new way to play music on the go. Most current methods involve manual music processing, using wearables, audio classification. Instead, it is recommended switching to manual connection and working traditionally. There convolutional neural network is used for emotion recognition. Using Pygame and Tkinter for better visualization. Our suggested approach will increase the system's overall efficiency by reducing the amount of time needed to get results and the overall cost of installation. Face expressions are captured by the built-in camera during system evaluation experiments on the FER2013 dataset. Concept face images are subjected to feature extraction in order to identify various emotions, including happy, angry, sad, surprised, and moderate. A user's current preferences are analyzed to update the music playlist. When compared to the algorithm in the current literature, it performs better in terms of computation time.