Telling Speaker Source apart in Speech to Text

Question

I am trying to get the transcript of a real-time video chat conversation happening over WebRTC on Web browsers. While I can get the transcript of the conversation with the likes