Music Genre Classification: A Comparative Analysis of CNN and XGBoost Approaches with Mel-frequency cepstral coefficients and Mel Spectrograms

TL;DR

This study compares CNN, VGG16, and XGBoost models using Mel spectrograms and MFCCs for music genre classification, finding that MFCC with XGBoost performs best and data segmentation improves CNN results.

Contribution

It provides a comparative analysis of deep learning and machine learning models with different audio features for music genre classification, highlighting the effectiveness of MFCC with XGBoost.

Findings

MFCC XGBoost outperforms CNN and VGG16 models.

Data segmentation significantly improves CNN performance.

MFCC features yield better results than Mel spectrograms.

Abstract

In recent years, various well-designed algorithms have empowered music platforms to provide content based on one's preferences. Music genres are defined through various aspects, including acoustic features and cultural considerations. Music genre classification works well with content-based filtering, which recommends content based on music similarity to users. Given a considerable dataset, one premise is automatic annotation using machine learning or deep learning methods that can effectively classify audio files. The effectiveness of systems largely depends on feature and model selection, as different architectures and features can facilitate each other and yield different results. In this study, we conduct a comparative study investigating the performances of three models: a proposed convolutional neural network (CNN), the VGG16 with fully connected layers (FC), and an eXtreme Gradient Boosting (XGBoost) approach on different features: 30-second Mel spectrogram and 3-second Mel-frequency cepstral coefficients (MFCCs). The results show that the MFCC XGBoost model outperformed the others. Furthermore, applying data segmentation in the data preprocessing phase can significantly enhance the performance of the CNNs.