Emotions pose a challenge when it comes to precise definition and interpretation. As a remedy, music emerges as a suggested art form for its potential in aiding mood regulation. Our proposed approach exhibits the potential to enhance system accuracy. This is achieved through a reduction in computation time for result generation, leading to improved overall efficiency and lowered operational costs of the devised system. To evaluate the effectiveness of our approach, we employed the FER2013 dataset, utilizing an integrated camera to capture facial expressions. The process involves extracting features from facial photographs to discern various emotions, encompassing happiness, anger, sadness, surprise, and neutrality. Leveraging a Convolutional Neural Network (CNN), we implemented a mechanism for detecting facial expressions, attaining an impressive detection accuracy rate of approximately 95.14 percent. The implementation employs a range of Python packages including WebRTC, Keras, and MediaPipe to facilitate the capture and analysis of users’ facial emotional cues.