Trying to transcribe sip conversation to text


I have successfully configured and have two working chan-sip users which use WEBRTC and sipml5 to communicate with each other. One user(100) is logged in the web browser and another user(200) in mobile on Zoiper. I have a paid service for Microsoft bing speech to text module, using it I want to save the conversation between the two speakers (no need to differentiate) in text. Added to it, save the audio recording either in mp3 or wav format.

Can anyone help, how to do it ?

Record the conversation, when the conversation is finished convert the recording to mp3 or wav and upload to your speech to text provider…

Would be a real challenge when you want the transcription be created realtime…

I do transcription using google speech recogntion API, I have done it also with microsoft but google it is more accurate, it is very simple just create the script using their API and you preferred programming language and pass the audio file as parameter