VAE-Model

VAE (Variational Autoencoders): These models focus on compressing images

into a compact "Latent Space" and then reconstructing them. They are

mathematically grounded but often produce slightly blurry outputs.