azure-stt-example

Start the container

To start the container, enter:

docker run --rm -it -p 5000:5000 --memory 4g --cpus 4 \
mcr.microsoft.com/azure-cognitive-services/speechservices/speech-to-text \
Eula=accept \
BILLING=<azure billing url> \
APIKEY=<api key>

Run the script

To only transcribe the firs word run:

python main.py --mode once

To transcribe the whole audio file:

python main.py --mode continuous

If you only want to see complete utterances without the intermediate results:

python main.py --mode continuous --show utterance

Audio file

Per default the file mary_and_sarah.wav will be transcribed, if you want to use another file instead, place it in the data-folder and specify it as an argument, like so:

python main.py --mode once --file mary_and_sarah.wav