R-GAN: Exploring Human-like Way for Reasonable Text-to-Image Synthesis via Generative Adversarial NetworksPublished in ACM MM, 2020Share on Twitter Facebook Google+ LinkedIn Previous Next