Hacker News new | ask | show | jobs
by nemo1618 3349 days ago
Clearly the solution is to employ a GAN, so that we simultaneously get artificial voices that are indistinguishable to the human ear, as well as judges that are able to reliably distinguish them.