README.md

imGANerator

• Domain transformation using Generative Adversarial Networks to generate images given a text description
  
• To conduct a thorough study of the codebase for the paper https://arxiv.org/pdf/1711.10485.pdf for Attentional Generative Adversarial Networks
  
• AttnGAN to generate images from caption inputs, encoded and inferenced over fine-grained classes; trained on MS-COCO dataset of annotated images to learn things, stuff, panoptics

Citations

@article{Tao18attngan,
  author    = {Tao Xu, Pengchuan Zhang, Qiuyuan Huang, Han Zhang, Zhe Gan, Xiaolei Huang, Xiaodong He},
  title     = {AttnGAN: Fine-Grained Text to Image Generation with Attentional Generative Adversarial Networks},
  Year = {2018},
  booktitle = {{CVPR}}
}