Audio Transcriber

This is a Python script to transcribe audio using the Google Speech API.

Prerequisites

This script requires you to have a Google Cloud project with the Speech API enabled (which requires billing to be enabled).

• Set up Google Cloud project and access to that project for users of this script.
• Make sure billing is enabled.
• Activate the Google Cloud Speech-to-Text API for the project.
• Create a storage bucket for temporarily holding audio files; put its name in the script

You must also have the Google Cloud command-line tools set up on the machine you will use to run this script.

• Install the gcloud command line tools following the steps in the official documentation
• Follow the steps all the way through running gcloud init.
  • When asked for a default project, use the one just created above.
• Log in to Google Cloud: gcloud auth application-default login

Installation

Install Python and the audio-conversion tool FFMpeg:

brew install python
brew install ffmpeg

Clone this repository and cd into its directory.

Create virtual environment for python:

/opt/homebrew/bin/python3 -m venv venv
source venv/bin/activate

And install the required Python packages:

pip install -r requirements.txt

Usage

cd to the application directory

Make sure the virtual environment has been activated:

source venv/bin/activate

And run the script, passing the path to the audio file you want to transcribe as a command-line argument:

./transcribe.py ~/Documents/my-audio.m4a

License

This project is (c) The Concord Consortium and licensed under the MIT License.