- Combine BicycleGAN(idea of using latent classifier as style encoder) and AttGAN
- Way to translate input image with disentangled latent input
- Most code brought from AttGAN and EigenGAN
- edges2shoes dataset is used
- epoch 160 trained model output (img size 128, z_dims 13)
- first col / second col / remian col
- ori source img / ori target img / output results when set value 4 to each input latent
- bicycleAttGAN.ipynb
-
Environment
-
Python 3.6
-
TensorFlow 1.15
-
OpenCV, scikit-image, tqdm, oyaml
-
we recommend Anaconda or Miniconda, then you can create the environment with commands below
conda create -n EigenGAN python=3.6 source activate EigenGAN conda install opencv scikit-image tqdm tensorflow-gpu=1.15 conda install -c conda-forge oyaml
-
NOTICE: if you create a new conda environment, remember to activate it before any other command
source activate EigenGAN
-
- ResNet
- Vae
- random z input
- tune hyperparameter
- test quality using Fréchet Inception Distance
- test diversity using lpips