danbamikiya / pit_transcriptor_google

Base Engine for Google Transcription

PIT Transcriptor

PIT Transcriptor utilizes the google cloud platform (specifically Google Speech to Text, and Google Storage) to create transcripts of long form audio.

Currently the process requires splitting audio into individual sections based on the where it returns a reliable split between speakers.

This project is in the very early stages of development.

Coming Eventually

• Support for Amazon Transcribe and other transcription tools
• Support for multichannel recognition
• Argument parsing via argparse
• The ability to upload files before processing them
• Asynchronous audio parsing
• A Web Front End that Allows people to Upload files from a web interface
• Tests (Those sooner than later)
• Auto-Selection of Transcription Algorithms based on Filetype
  • WAV/Flac - Google Speech to Text
  • MP3 - Amazon Transcribe

Dependencies

• Google Cloud Platform account
• A JSON Credential (Google Cloud) with the path saved to your environments
• Python 3.5+ (Becuase of f'string and {such}')
• Pipenv for package/moduled/dependency management

Contributions

I'm not currently accepting contributions at this time.