Hi all,
I just saw these two video demonstrations about audio noise cancelation/reduction. It looks much like those noise cancelation features from Bose and Sony headphones, but all through the use of machine learning.

1. Stationary vs non-stationary noises with Krisp App
2. 2Hz demo

I want to do something like this on my own. Can someone give me more information on the (most likely) used neural network and overall setup to achieve audio noise cancelation/reduction like in the videos?
Thanks in advance!