AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
How to Diarize Speakers in Python
This is something that we very much desire for our transcripts because we have it already with our human transcribed transcripts. There were like few ideas popping around using whisper in no traditional way to achieve some sort of diarization. It boils down to trying to extract some of the internal results of the computation and try to like classify based on some, let's say, features or I'm not sure really how to properly explain it. For now it seems unlike it. It's a pipe dream. This slow unlikely, at least from my point of view,. maybe if someone fears it out and it pretty works, we could probably have it someday.