EatMyGoose / Omnizart-UI

Browser based UI for music transcription tool Omnizart

Omnizart UI

About

Omnizart UI is a browser-based UI to simplify use of Omnizart - a ML music transcription tool that converts music files into MIDI files.

Usage Guide

Installation

• Prerequisites:
  • Docker Desktop installed
• Clone this repository into a folder of your choosing
• Set the working directory to the folder containing the docker-compose.yml file
• Run docker compose up in the terminal
• Navigate to http://localhost:5173/

Converting a music file to MIDI

1. Upload a file within the Select Input File section
2. Select the desired transcription mode
   • Music (Transcribes all instruments/vocals)
   • Vocal (Extracts out only the melody)
3. Click Transcribe
   • Alternatively, toggle Auto-Transcribe to On to start the transcription process immediately after uploading a file
4. Wait for transcription to finish (this may take a while)
5. To hear a preview of the transcribed file, start the player in the Transcription Result section
6. Click on Download to download the resultant MIDI file.

Downloading a previously transcribed file

1. Navigate to the sidebar, click the › button to expand it if neccessary
2. Find the transcribed file and click on the download icon
   • Results can be filtered based on filename in the Settings box

Remarks

• Omnizart UI not directly affliated with the original creators of Omnizart
• The transcription history is stored in non-persistent storage, destroying the backend container will erase the transcription history
• Modes Drums & Chords are currently unsupported due to Python version dependency issues in Omnizart (Python 2.7 required)
• Vocal Contour is omitted, since it produces a .wav file instead of a midi transcription.

Attributions

Citation as requested by the authors of Omnizart:

[Omnizart]
@article{Wu2021,
  doi = {10.21105/joss.03391},
  url = {https://doi.org/10.21105/joss.03391},
  year = {2021},
  publisher = {The Open Journal},
  volume = {6},
  number = {68},
  pages = {3391},
  author = {Yu-Te Wu and Yin-Jyun Luo and Tsung-Ping Chen and I-Chieh Wei and Jui-Yang Hsu and Yi-Chin Chuang and Li Su},
  title = {Omnizart: A General Toolbox for Automatic Music Transcription},
  journal = {Journal of Open Source Software}
}