Hello,
Sapir and I(Haim) would like to present paper about generate realistic image from text using BERT + STACKGAN
link to the pdf: https://arxiv.org/pdf/1612.03242.pdf
In the paper there are text description of a bird for an example, and from its description we produce a composite image that tries as much as possible to simulate what is described in the description we received.
Hi Sapir and Haim, Approved.