Transcribe job with speaker identification can not be downloaded with speaker identification

Question

Hello, I have used the Transcribe tool, and made job  I have selected to identify the Speakers- I can see this it worked but only in the Transcription preview - there I can see Speaker 0 and Speaker 1, I also need to download the transcript, but when I do - in the JSON file I can see only Raw data without this identification. Could you please help me to pull the transcript with audio identification or is it actually in the json

Answer

Hope you are doing well.

To enable speaker identification, you need to enable the "Speaker identification" option.

By default, AWS Transcribe provides you the overall JSON file (response) with speaker information. To retrieve the text from respective speaker, you need to parse that JSON file. I have written/covered the same in one of my video [here](https://youtu.be/4qGmRgO2WNE) along with the [script](https://github.com/srcecde/aws-tutorial-code/blob/master/lambda/lambda_transcribe_speaker_seperation.py)

I hope this helps :)

Transcribe job with speaker identification can not be downloaded with speaker identification

Contenus pertinents