Audio classification with transcription

I don't know if it's a good idea, but having an option to classify different sounds of an audio clip while doing the transcription. Or without doing any transcription.

I see this working that way:

1) You select an option to activate the audio classification. The first pass transcribes the audio; the second pass tries to classify "sounds—cheering, applause, laughs, car engines, etc.—and . adds them somewhere in the transcription window.

Maybe Resolve could use its own dialogue separator for these tasks.