ViewTube

... especially variations of neural networks such as Generative Adversarial Networks (GANs) or Variational Autoencoders (VAEs).

... approach in which a Vision Transformer processes compressed image latents from a frozen Variational Autoencoder to produce ...