Exploring Machine Learning Techniques for Music Emotion Classification: A Comprehensive Review

Music Emotion Recognition (MER) has made substantial advancements in recent years, driven by improvements in signal processing, machine learning, and interdisciplinary collaborations. The work analyzes the most recent advancements in MER, with a specific emphasis on fundamental methodologies, obstacles, and emerging patterns. The review commences by clarifying the fundamental theories of music and emotion, emphasizing the complex interaction between auditory stimuli and emotional reactions. The text examines the progression of feature extraction methods from conventional acoustic features to deep learning representations, offering a comprehensive insight into the techniques used to capture the subtle emotional nuances present in music signals. The paper thoroughly examines the various machine learning models used in MER, emphasizing their advantages and drawbacks. Further, the review delves into the accessibility and variety of datasets, recognizing their crucial importance in training resilient models for emotion recognition. At last, it provides a thorough analysis of the current state of Music Emotion Recognition, with the goal of assisting researchers, practitioners, and enthusiasts in exploring the ever-changing field. It also aims to encourage further progress in comprehending the deep relationship between music and human emotions.