pix2pix

Keras implementation of Image-to-Image Translation with Conditional Adversarial Networks by Phillip Isola, Jun-Yan Zhu, Tinghui Zhou, Alexei A. Efros

Requirements

python modules

• numpy==1.13.3
• Keras==2.0.8
• parmap==1.5.1
• matplotlib==2.0.2
• tqdm==4.17.0
• opencv_python==3.3.0.10
• h5py==2.7.0
• theano==0.9.0 or tensorflow==1.3.0

Part 1. Processing the data

Follow these instructions.

Part 2. Running the code

Follow these instructions

Part 3. Example results

Facade example results (validation set)

For each image:

• The first row is the generator input
• The second row is the generator output
• The last row is the original image