Voice Beautifier

While Discord offers filters for noise, it would be very nice to also have an option to beautify your voice (e.g. compression, anti-hissing, reverb, equalization, removal of noise, etc.) using artificial intelligence.

The procedure should be fairly straight forward to implement using deep learning. As training material, I would suggest raw data from speech recordings as input and the edited counterparts as ground truths. I guess that such data sets are either already available or could easily be attained from recording studios that specialize on audio books and stuff.