Any way to add speaker diarization feature to this code base?

[RYucel]

Due to some small project in my mind, recently I involved in speaker diarization but could not find any suitable method except python implementation. Is there any way to add in this project a speaker diarization option for faster subtitling than other methods (c and c++ have huge potential I guess) ?

[ggerganov]

Stereo-channel-based diarization will be added soon - see #64 for more info.
It is simple to implement and I hope it will work well enough for the cases of 2 people speaking with stereo audio.

More complex algorithms are unlikely to be added to this project because it seems the general problem of diarization is very complex.