Classification from Raw Audio, Mel spectrogram features, and STFT Images are done in this repo.
Explore the docs »
View Demo
.
Report Bug
.
Request Feature
In this repo,
- You will find tow Jupyter Notebook files, one you can use as a step-by-step guide and other is for STFT Transformation from Audio waveforms. Every step is explained thoroughly in the notebooks.
In detail, the performance of three different methods for music genre classification using the FMA small version dataset is explored. Firstly, the use of raw audio waveforms as features for classification is investigated. Secondly, mel spectrogram features, which are commonly used in music information retrieval tasks, are used for genre classification. Finally, the performance of classification using Short-Time Fourier Transform (STFT) images, which provide a compact representation of audio signals, is evaluated. The aim of this study is to compare and evaluate the performance of these three methods for music genre classification and to provide insights into the most suitable approach for this task. The main contributions of this project are:
- A comparative study of three different music genre classification methods: raw audio waveform, mel spectrogram features, and STFT images.
- A thorough evaluation of the three methods on the FMA small dataset, comparing the results to previous work in the literature.
- An analysis of the results and demonstration of the effectiveness of the proposed methods.
This Jupyter Notebook was written in Python & Pytorch. You can directly download the notebook and run it locally, or you can upload it on Google Collabs and run there (I recommend GPU usage for faster training time).
To get a local copy up and running follow these simple example steps.
Clone the repo
git clone https://github.com/mustafaAlgun/Audio-Classification-PyTorch.gitDataset can be downloaded from this repo.
This is a walk-through Jupyter Notebook. Therefore, following the steps and reading comments will be more than enough for you to replicate the results :) Mind you, you may need to change the paths of files according to where they are in your local machine.
See the open issues for a list of proposed features (and known issues).
Contributions are what make the open source community such an amazing place to be learn, inspire, and create. Any contributions you make are greatly appreciated.
- If you have suggestions for adding or removing projects, feel free to open an issue to discuss it, or directly create a pull request after you edit the README.md file with necessary changes.
- Please make sure you check your spelling and grammar.
- Create individual PR for each suggestion.
- Mustafa Algun - Master's Student at University of Padua
@inproceedings{fma_dataset, title = {{FMA}: A Dataset for Music Analysis}, author = {Defferrard, Micha"el and Benzi, Kirell and Vandergheynst, Pierre and Bresson, Xavier}, booktitle = {18th International Society for Music Information Retrieval Conference (ISMIR)}, year = {2017}, archiveprefix = {arXiv}, eprint = {1612.01840}, url = {https://arxiv.org/abs/1612.01840}, }