A Style-Based Generator Architecture for Generative Adversarial Networks
IEEE Transactions on Pattern Analysis and Machine Intelligence(2021)
摘要
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-level attributes (e.g., pose and identity when trained on human faces) and stochastic variation in the generated images (e.g., freckles, hair), and it enables intuitive, scale-specific...
更多查看译文
关键词
Generators,Convolution,Training,Image resolution,Aerospace electronics,Generative adversarial networks,Interpolation
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要