Using Residual Generative Adversarial Networks and Variational Auto-encoder techniques to produce high resolution images.
Python
Switch branches/tags
Nothing to show
Clone or download
HaDavid HaDavid
HaDavid and HaDavid added video
Latest commit 9206e06 Jun 2, 2016
Permalink
Failed to load latest commit information.
MNIST_data first commit Jun 2, 2016
examples added video Jun 2, 2016
save first commit Jun 2, 2016
README.md first commit Jun 2, 2016
images2gif.py first commit Jun 2, 2016
mnist_data.py first commit Jun 2, 2016
model.py first commit Jun 2, 2016
ops.py first commit Jun 2, 2016
sampler.py first commit Jun 2, 2016
train.py first commit Jun 2, 2016

README.md

resnet-cppn-gan-tensorflow

Improvements made for training Compositional Pattern Producing Network as a Generative Model, using Residual Generative Adversarial Networks and Variational Autoencoder techniques to produce high resolution images.

Morphing2

Run python train.py from the command line to train from scratch and experiment with different settings.

sampler.py can be used inside IPython to interactively see results from the models being trained.

See my blog post at blog.otoro.net for more details.

Morphing1

I tested the implementation on TensorFlow 0.80.

Used images2gif.py written by Almar Klein, Ant1, Marius van Voorden.

License

BSD - images2gif.py

MIT - everything else