VAE: variational autoencoders and the latent space
You cannot generate a new image by making up pixel values at random — the space of all possible pixel combinations is astronomically large, and almost none of them look like anything meaningful. Generative models work by learning a much smaller, structured latent space where every point corresponds to a plausible image. The VAE is the first architecture that made this idea practical, and its latent space is the foundation of Stable Diffusion.
Content is available with subscription.
Get full access to all courses on the platform for one year with a single payment.
▼
Unlike other platforms that charge per course, here you get everything for one price, and after one year of use there will be no automatic charge for the following year.