In our research, we fine-tuned the DistilHuBERT model using the GTZAN dataset for music genre classification. We evaluated the effectiveness of three different data augmentation techniques.

Introduction: This repository contains code for a music genre classification project using machine learning techniques. The project aims to classify audio clips into various music genres, providing a valuable tool for music enthusiasts, streaming platforms, and researchers.

Project Structure:

• data/: This directory contains the GTZAN dataset, which comprises audio clips from ten different music genres.
• models/: This directory contains pre-trained models and fine-tuned models used for audio classification.
• utils/: This directory contains utility functions for data preprocessing, model evaluation, and visualization.
• train.py: This script is used to train and fine-tune the classification model on the GTZAN dataset.
• evaluate.py: This script is used to evaluate the performance of the trained model on a separate test set.
• predict.py: This script allows users to classify audio files using a trained model.

Getting Started:

1. Clone the repository to your local machine: git clone <repository-url>
2. Install the required dependencies: pip install -r requirements.txt
3. Download the GTZAN dataset and place it in the data/ directory.
4. Train the classification model using the train.py script.
5. Evaluate the model's performance using the evaluate.py script.
6. Use the predict.py script to classify audio files.

Usage:

• train.py: Run python train.py to train the classification model. Adjust the training parameters as needed in the script.
• evaluate.py: Run python evaluate.py to evaluate the trained model on the test set.
• predict.py: Run python predict.py <audio-file> to classify a single audio file using the trained model.

Contributing: Contributions to this project are welcome. If you encounter any bugs or have suggestions for improvements, please open an issue or submit a pull request on GitHub.

Acknowledgments:

• The GTZAN dataset used in this project was originally created by George Tzanetakis and Perry Cook and is widely used for music genre classification research.
• We would like to thank the creators of the Hugging Face Transformers library for providing pre-trained models and tools for natural language processing and audio classification tasks.