In this paper, we focus on generating realistic images from text descriptions. Current methods first generate an initial image with rough shape and color, and then refine the initial image to a ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results