/AttnGAN-Tensorflow

Simple Tensorflow implementation of "AttnGAN: Fine-Grained Text to Image Generation with Attentional Generative Adversarial Networks" (CVPR 2018)

Primary LanguagePythonMIT LicenseMIT

AttnGAN — Simple TensorFlow Implementation [Paper]

: Fine-Grained Text to Image Generation with Attentional Generative Adversarial Networks

TO DO

  • Deep Attentional Multimodal Similarity Model (DAMSM)

Dataset

Text

Image

Usage

├── dataset
   └── YOUR_DATASET_NAME
       ├── images
           ├── domain1 (domain folder)
               ├── xxx.jpg (domain1 image)
               ├── yyy.png
               ├── ...
           ├── domain2
               ├── aaa.jpg (domain2 image)
               ├── bbb.png
               ├── ...
           ├── domain3
           ├── ...
       ├── text
           ├── captions.pickle
           ├── filenames_train.pickle
           ├── filenames_test.pickle

Train

python main.py --dataset birds --phase train

Test

python main.py --dataset birds --phase test

Author

Junho Kim