To begin:
git clone https://github.com/justindixon/AudioTranscription.git
cd /AudioTranscription
This is a simple program that uses the Google Cloud Natural Language API to transcribe audio files.
You must use your own API key which can be obtained by first registering for Goolge Cloud Platform:
https://console.cloud.google.com/start
https://cloud.google.com/billing/docs/how-to/modify-project#enable_billing_for_a_project
Once you have completed that then you need to log into the console and enable the Cloud Speech API. Once enabled then you need to obtain an API KEY:
https://cloud.google.com/docs/authentication/api-keys#creating_an_api_key
https://cloud.google.com/video-intelligence/docs/common/auth
Once you have that key you need to store the file in the directory and enter this command in the terminal:
export GOOGLE_APPLICATION_CREDENTIALS=<path_to_service_account_file>
To install the program please enter this command:
pip install -r /requirements.txt
This will install the python modules required for the app.
You may also need to install:
- ffmpeg
- sox
Once that is all set up you can then use the program as such:
python app.y --audioFile=<path_to_audio_file_here>
In addition you can add:
--sentiment=True
This will return, in addition to the transcript, the sentiment of the audio file.
Future developments:
- cost estimation with option to continue
- ability to integrate a grammar checker for better output.
- create access to more Google Cloud Natural Language API options
- Cleaner output
- test module