R-GAN: Exploring Human-like Way for Reasonable Text-to-Image Sythesis via Generative Adversarial Networks

Publication
ACM MM