Multiple speakers within one record for training

[JRMeyer]

>>> vcjobacc
[September 11, 2019, 7:27am]

Hello everyone! slash
Just wanted to find out how bad would it be to use records which have
more than one speaker talking. slash
Let me explain a bit more. I extract speech data from youtube based on
manual subtitles provided. To collect as much data as possible within a
short time period I perform almost no post processing. The music, noise
and other kind of acoustic effects are kept - I guess and hope it will
lead to more robust model. Am I wrong? slash
And as long as subtitles has no info regarding the speakers (who spoke
when) and I leave it as it is, quite often there are multiple people's
speech being presented within a single record. How bad it is (if is)?
After all I'm gonna add this data to a clean dataset of 300 Hours (with
single speaker per single record). slash
Thank you all for the suggestions!

[This is an archived TTS discussion thread from discourse.mozilla.org/t/multiple-speakers-within-one-record-for-training]

[JRMeyer]

>>> tanmayjain
[December 4, 2020, 10:59am]

Hi, slash
Sorry for bump last year's topic. slash
Did you try this? slash
Did multiple speakers in a single chunk of audio effects the result?

Thanks

[Archived Post]

[JRMeyer]

>>> othiele
[December 4, 2020, 12:50pm]

We didn't have any problems with multiple speakers in one chunk as long
as they don't speak at the same time.

[Archived Post]