Music Genre Classification Using Convolutional Neural Networks

Authors: Siddhartha T

Abstract: Automatically classifying music into genres is a challenging task that has seen significant improvement through the use of deep learning. In this paper, we present a Convolutional Neural Network (CNN)-based model for music genre classifi- cation using Mel-Frequency Cepstral Coefficients (MFCCs) ex- tracted from audio files. Our model was trained and evaluated on the GTZAN dataset, achieving a solid accuracy of 91.2%. These results highlight the potential of deep learning to understand and categorize audio content effectively.