Paper Title
MUSIC GENRE CLASSIFICATION USING SEMI-SUPERVISED GENERATIVE ADVERSARIAL NETWORK (SGAN)

Abstract
With the advent of music streaming, the music industry has undergone a significant shift. To cater to the tastes of music consumers, there is now a growing need and interest in developing automatic music genre classification systems. This would enable the selection and recommendation of music based on the listener's preferences, helping to ensure a better user experience. As a result, there is currently active research and demand for such systems in the music industry. One of the most prevalent characteristics used to categorize musical works is genre. Despite the existence of various genre definitions worldwide, human biases can affect how people respond to different genres of music. By analyzing the transition of music to digital platforms, it becomes apparent that automating music classification could be advantageous for numerous fields. To that end, this study investigates genre classification using a generative adversarial network (GAN) on wave images of GTZAN dataset's test data. Keywords - GAN, Machine Learning, Semi-Supervised Learning, Spectrogram, Music Genre.