This repo shows how to implement a real time transcription using Azure AI Speech service. We will be using the Python SDK. 4 different scripts have been implemented making use of a common helper module. The scenarios covered are:

• Transcription from an MP3 audio file
• Transcription from an MP4 video file
• Transcription from a WAV audio file
• Transcription from a WAV audio input stream

Create an .env file or enable the following values as system variables:

speech_key="YOUR_SPEECH_SERVICE_KEY"
speech_region="YOUR_SPEECH_SERVICE_REGION"
speech_language="en-us" #YOUR DESIRE LANGUAGE

Install packages and depencies with the following script

./aux_installer.sh

Load video content in the media folder using the following script. The source is defined in the config.sh file in the input variable.

./aux_content_generator.sh

A use case for each of the 4 scenarios listed above are implemented in the script aux_tester.sh

./aux_tester.sh

Testing MP3 file
Transcribing media/input.mp3 
SESSION STARTED
RECOGNIZING: if you're
RECOGNIZING: if you're a
.......