Keras implementation of Deep Convolutional Generative Adversarial Networks
Switch branches/tags
Nothing to show
Clone or download
Latest commit 0603639 Jul 21, 2017
jacobgil committed Jul 21, 2017 Merge pull request #17 from jusjusjus/master
Updated keras commands for v2.0+.
Failed to load latest commit information.
assets readme assets Aug 26, 2016 Add theano requirement to readme Nov 9, 2016 Updated keras commands for v2.0+. Jul 19, 2017


Implementation of with the (awesome) keras library, for generating artificial images with deep learning.

This trains two adversarial deep learning models on real images, in order to produce artificial images that look real.

The generator model tries to produce images that look real and get a high score from the discriminator.

The discriminator model tries to tell apart between real images and artificial images from the generator.

This assumes theano ordering. You can still use this with tensorflow, by setting "image_dim_ordering": "th" in ~/.keras/keras.json (although this will be slower).



python --mode train --batch_size <batch_size>

python --mode train --path ~/images --batch_size 128

Image generation:

python --mode generate --batch_size <batch_size>

python --mode generate --batch_size <batch_size> --nice : top 5% images according to discriminator

python --mode generate --batch_size 128


generated images :



train process :