Music Genre Classification using Machine Learning on FMA Dataset

Abstract

Music is a universal form of expression with a multitude of genres that resonate with diverse audiences. While genre classification may seem straightforward to the human ear, automating this process poses a complex challenge. This complexity stems from the subtle and intricate characteristics that differentiate one musical genre from another. Effective categorization not only has implications for how music is organized and recommended in digital platforms but can also provide insights into the underlying structure and semantics of musical compositions. To tackle this issue, we aim to utilize machine learning and deep learning techniques to automatically categorize music into genres.