Hybrid Deep Learning Model for Mood-Driven Song Recommendations

Human experiences are greatly influenced by emotions, and music has long been understood to be a potent medium for influencing, reflecting, and elevating emotional states. This study introduces a novel system that integrates facial expression recognition with a music recommendation engine to create a dynamic, emotion-driven music experience. The system detects subtle emotional cues by accurately extracting key characteristics from facial photos using Convolutional Neural Networks (CNNs) and attention mechanisms. The incorporation of the Convolutional Block Attention Mechanism (CBAM) enhances the model’s ability to recognize complicated facial expressions by capturing global contextual information. A personalized music playlist is created by querying the Spotify API after the emotion has been identified and mapped to a matching mood category. This work demonstrates how merging cutting-edge AI algorithms can result in a more engaging and intuitive user experience motivated by emotional understanding.