Museko is a music analysis tool.

• Genre classification with Essentia TensorFlow models (discogs EffNet CNN, genre_discogs400)
• YouTube mp3 downloader (Python: yt-dlp, Flask)
• Real-time audio visualizer (JavaScript: audioMotion-analyzer)
• BPM and additional song details (Python: librosa, essentia)

Usage

Unfortunately, this tool requires too much memory to be hosted online using free services. The only way to try out this app would be running it locally.

Instructions

Note: essentia-tensorflow is not supported for Windows.

1. Clone and navigate to the repository:

git clone https://github.com/bmai1/Museko.git
cd Museko

2. Install Python dependencies. It is recommended to use a virtual environment:

python -m venv env
. env/bin/activate

pip install -r requirements.txt

3. Navigate to the inner museko folder and run:

cd museko
flask run

4. Flask development server defaults to http://127.0.0.1:5000, which you can open in a web browser

5. Upload your mp3 files and wait for the genre classification model to analyze it, after which it will display the genre graph and audio visualizer

6. Ctrl-C to close the server when you are done