How to anonymize a voice recording? How to hide/remove noise?

Question

Algorithms

Faced such a problem: there is a .wav (if it is important) file with a voice that needs to be processed to the degree of indeterminacy of gender / age.
Second: the audio file is compressed very strongly, then decompressed (through the Fourier transform), the recording quality drops greatly. What are the ways to remove interference?

Interested in open projects on these topics, algorithms for solving these problems.

Answer 1

Convert voice to text Convert
text to voice with any available speaker

Answer 2

to the degree of indeterminacy of sex/age

What if the person says "I did" or "I did"?
And if he says "calling" instead of "calling"?
What if he ok or kat?
And if he uses words that are characteristic of a certain group of people ("front door", "curb", "shawarma" or "auf", "padra", "vislovo")?

Answer 3

You can change the playback speed and pitch in the audio editor (only preserve one of them when you change the second) and play around with the equalizer. There will be anonymization, but with less loss of sound quality.
Tones and accents will still be preserved, so if you need direct complete anonymization, then, as already said, it is better to use a speech synthesizer.