Hacker News new | ask | show | jobs
by spywaregorilla 1618 days ago
Does the discriminator model translate the images into an embedding space of it's own? Could such a space be used to generate images themselves?