Unsupervised Projection Networks for Generative Adversarial Networks

09/30/2019

∙

We propose the use of unsupervised learning to train projection networks that project onto the latent space of an already trained generator. We apply our method to a trained StyleGAN, and use our projection network to perform image super-resolution and clustering of images into semantically identifiable groups.

READ FULL TEXT

Unsupervised Projection Networks for Generative Adversarial Networks

ClusterGAN : Latent Space Clustering in Generative Adversarial Networks

Optimizing Generative Adversarial Networks for Image Super Resolution via Latent Space Regularization

Learning to Associate Words and Images Using a Large-scale Graph

Fine-grained Attention and Feature-sharing Generative Adversarial Networks for Single Image Super-Resolution

An Image-based Generator Architecture for Synthetic Image Refinement

Projection predictive variable selection for discrete response families with finite support

Efficient Unsupervised Learning for Plankton Images

Unsupervised Projection Networks for Generative Adversarial Networks

Related Research

ClusterGAN : Latent Space Clustering in Generative Adversarial Networks

Optimizing Generative Adversarial Networks for Image Super Resolution via Latent Space Regularization

Learning to Associate Words and Images Using a Large-scale Graph

Fine-grained Attention and Feature-sharing Generative Adversarial Networks for Single Image Super-Resolution

An Image-based Generator Architecture for Synthetic Image Refinement

Projection predictive variable selection for discrete response families with finite support

Efficient Unsupervised Learning for Plankton Images