Autoencoders
Stacked Autoencoder
Autoencoders Types
Unsupervised Pretraining Using Stacked Autoencoders
Denoising Autoencoders
Variational Autoencoders
Generative Adversarial Networks (GANs)
GAN Training Algorithm
Difficulties of Training GANs
StyleGANs
State of the Art
Use style transfer techniques in the generator to ensure that the generated images have the same local structure as the training images, at every scale, greatly improving the quality of the generated images. The discriminator and the loss function were not modified, only the generator
GAN Style Transfer
Diffusion Models
Zuletzt geändertvor 7 Monaten
