Include generation and manipulation
guided 有一种引导的意思在里面，所以生成和操作不一样，所以要分开，最新的文章是将这两者结合起来，
Key-words: Text-to-Image Generation, Text-Guided Image Manipulation
也有人写 Text-Guided Image Generation
希望首先有一句话的概括，然后是文献的具体，引用缩写可以用bibtex的缩写
用户自己的话
(StyleCLIP)-Text-Driven Manipulation of StyleGAN Imagery
[Arxiv 2021] (TAU) Or Patashnik, Zongze Wu, Eli Shechtman, Daniel Cohen-Or, Dani Lischinski
main method: combine recurrent neural networks and generative adversarial networks
Learning What and Where to Draw
Generative adversarial text-to-image synthesis
Parallel Multiscale Autoregressive Density Estimation
StackGAN: Text to Photo-realistic Image Synthesis with Stacked Generative Adversarial Networks