# 🎧🎵 Boost Global Music Genre Classification — Data

来源：https://www.kaggle.com/competitions/boost-global-music-genre-classification/data

本地 Meta Kaggle 快照；不是本次实时页面。

# Dataset Description

📂 Dataset: MuGen-100 – A Global Music Genre Classification Dataset:

MuGen-100 is a comprehensive dataset designed to advance music genre classification by encompassing a diverse array of musical genres from around the world. It includes 100 distinct genres, ranging from widely recognized categories like Pop and Classical to region-specific styles such as Ahidous, Amapiano, and Guqin.

📁 Dataset Contents:

- Size: 43.5 GB

- 100 Genres: A curated selection representing various cultures and musical heritages.

- Audio Samples: Each genre is represented by 100 audio clips of 30 seconds, providing a rich dataset for training and evaluation.

- Feature Representations: (Not included) Pre-extracted features such as Mel-Frequency Cepstral Coefficients (MFCCs) and spectrograms to facilitate model development. 
But we used the pre-processing steps shown here: (from a YouTube tutorial) to get them: https://github.com/animesh1012/machineLearning/blob/main/Music_Genre_Classification/Train_Music_Genre_Classifier.ipynb

🔗 Accessing the Dataset

The dataset is hosted on SharePoint and can be accessed via the following link on Github.


📄 Additional Information
For detailed information about the dataset, including its sources, creation, and structure, refer to the accompanying documentation on Github.