Advancing Music Genre Identification Through Deep Learning Techniques

Music plays a huge role in human's lives, inspiring a wide range of feelings including enthusiasm and nostalgia. 97 million songs have been recorded globally; an incredible figure that is only increasing exponentially as more people use music applications. Many people love particular musical genres, such as classical, hip-hop, or disco, and want for an easy way to arrange their music in accordance with their preferences. This gave rise to the idea of “music genre classification”. In the current environment, where there are many music recordings available both online and offline, proper genre categorization has become essential. Several machine learning methods may be used to categorize the different types of music. This model's emphasis is on using deep learning methods to identify the musical genre. The Mel-frequency cepstral coefficients (MFCC), a technique that works exceptionally well for music files, are used to extract key elements. The proposed model is excellent at categorizing musical subgenres and provides default settings as well as interactive options depending on user preferences. The simple web interface built using Flask allows users to submit audio files and view the anticipated genre. The system also has the capability of recommending new genres to the user.