A Style-Based Generator Architecture for Generative Adversarial Networks

文章目录
Abstract1. Introduction2. Style-based generator2.1 Quality of generated images
3. Properties of the style-based generator3.1. Style mixing3.2. Stochastic variation

Abstract

We propose an alternative generator architecture for generative adversarial networks, borrowing from style transfer literature. The new architecture leads to an automatically learned, unsupervised separation of high

A Style-Based Generator Architecture for Generative Adversarial Networks最先出现在Python成神之路。