MeshTalk Realtime Implementation

Hello, I want to implement a major pipeline. I would like this https://github.com/facebookresearch/meshtalk to be implemented. I have also attached the paper of this implementation. I would to have Audio (from laptop microphone) to a text to speech program (any good text to speech program will do) and then to the meshtalk in REAL TIME (I want to talk into my microphone and have a model blue man like in the paper emulate it in real time). I would like that to be implemented, and I would also like the same thing implemented but using an automatic speech recognition software, so this would be audio (from laptop microphone) to text to speech to automatic speech recognition program to meshtalk in real time or vice versa, and I would like to see this implemented. I would like to know step-by-step how to do it (on your end, what I need to install and where), and examples of it.