Novel Music Genre Classification System Using Transfer Learning on a Small Dataset

Abstract

Music is an essential "element" in our lives. With the rapid development of technology, the use of compact disks and tapes for listening various forms of music of different cultures has become obsolete. Because of the generation of considerable music data, the accurate classification of music data has become a critical topic of research, and the development of deep learning technology provides a solution for music classification. Transfer learning techniques are widely used in image recognition but are rarely used for music classification. In this study, a small music dataset is used for music classification using transfer learning.