Check out
Spleeter from Deezer here :
https://github.com/deezer/spleeter.
It's mostly used to extract instruments and vocals. I personally used it on other things than songs... regular audio with voices and background noises.
It was a while back so I don't remember exactly, but I'm pretty sure that by selecting "2 stems", I got the voice in one file, and whatever was in the background in a second file.
On the github page, they say :
Spleeter pre-trained models have also been used by professional audio softwares. Here's a non-exhaustive list:
iZotope in its Music Rebalance feature within RX 8
SpectralLayers in its Unmix feature in SpectralLayers 7
Acon Digital within Acoustica 7
VirtualDJ in their stem isolation feature
Algoriddim in their NeuralMix and djayPRO app suite
So I imagine this type of operation can be done, if Resolve can isolate voices, it can isolate the background noise.
I didn't use any of the mentioned tools but iZotope is a standalone and a plugin for DAWs right?
So I imagine that the operation can be done in real time.