I expected something this: https://speech2face.github.io/ https://speech2face.github.io/ (arbitrary voices) .. this model seems to have been trained for each an

I expected something this: https://speech2face.github.io/ https://speech2face.github.io/ (arbitrary voices) .. this model seems to have been trained for each and every specific speaker?