Audiobeep

This is a full-stack audio modifier application built with ReactJs and Python. It uses the package such as Demucs and WhisperX for seperating songs into foreground & background and extracting the transcript from the foreground. Other techniques are then applied to censor profanities and provide a high quality output with unnoticeable censored points.

Features

• Multi upload and Processing: Uploading of multiple songs in one go, or over intervals with concurrent processing of uploaded contents.
• Real time voice profanities censoring: Real time profanities censoring with live streaming link for tuned in audience.
• Song Transcript reviewer: All songs processed have generated transcript which can be reviewed and readjustments can be made.
• Upload History: Paginated view of all songs processed available for reviewing and downloading of the processed audio.

Technologies

• ReactJs: A javascipt frontend framework for building static and dynamic websites.
• Python: A high level programming language with simple english like syntax.
• FastAPI: A python web framework which utilizes ASGI, that is fast ant efficient.

Other Technologies Used

• AWS, RecordJs, SocketIO.js and Pydub

Conclusion

You can check out the website via the link at Audiobeep