A Brief Survey of text driven image generation and maniulation

Text-to-image synthesis refers to the translation of textual description to images with similar semantic meaning. This process mainly relies on the correlation of text and image to find the best alignment between them. So, the recent development of generative models, especially GANs, using natural language as a condition paves a new path for research. In this paper, we review the most notable developments in the Field of text-to-image synthesis. In contrast to others, were viewed new methods of GAN-based text to image synthesis comparatively, considering two meaningful divisions, generation and manipulation. Additionally, there is a summary table for the characteristic of all these methods