How to recognize speech by separating voices in single channel audio?

Question

Speech recognition

Such a task, there are single-channel wav tracks recording a conversation between two people. It is necessary to translate the speech into text, but separating the owners of the voice. Now I use Yandex speechkit, but it translates all the voices into text all the way. How to split text by owners. automated? Not important . in what programming language.

Answer 1

Google speech could, it's called diarization, I'm attaching a link to an example, maybe it will help someone.
https://cloud.google.com/speech-to-text/docs/multi...