Music Genre Classification System Using Deep Learning Algorithm

The capacity to categorize music files into distinct genres is essential for improving user experiences and customizing suggestions in the age of digital music consumption. This work aims to accurately classify music files into different genres by providing a thorough analysis of deep learning algorithm-based music genre classification system. Utilizing deep learning and signal processing ideas, the system generates wavelets and spectrograms to transform audio inputs into formats that can be analyzed. The study shows that convolutional neural networks (CNNs) are an effective tool for processing and categorizing music files according to their unique genre attributes. Through rigorous experimentation and analysis, the research highlights the importance of accurate genre classification in increasing user engagement and enjoyment with digital music. The experiments are conducted on the GTZAN (Olteanu in GTZAN dataset—music genre classification, 2020) dataset, and we obtained an accuracy of 91% with the CNN model which is the maximum when compared with other models such as SVM, KNN, and random forest. We also have the values of some metrics for evaluation and how well the model undergoes training is depicted by the loss curve and the prediction accuracy by the accuracy curve.