first commit

Author: HaDavid

MNIST_data/t10k-images-idx3-ubyte.gz

@@ -0,0 +1,21 @@
+# cppn-gan-vae tensorflow
+
+Train [Compositional Pattern Producing Network](https://en.wikipedia.org/wiki/Compositional_pattern-producing_network) as a Generative Model, using Generative Adversarial Networks and Variational Autoencoder techniques to produce high resolution images.
+
+![Morphing](https://cdn.rawgit.com/hardmaru/cppn-gan-vae-tensorflow/master/examples/output_linear.gif)
+
+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](http://blog.otoro.net/2016/04/01/generating-abstract-patterns-with-tensorflow/) for more details.
+
+I tested the implementation on TensorFlow 0.60.
+
+Used images2gif.py written by Almar Klein, Ant1, Marius van Voorden.
+
+# License
+
+BSD - images2gif.py
+
+MIT - everything else

MNIST_data/t10k-labels-idx1-ubyte.gz

@@ -0,0 +1,30 @@
+'''
+generate cool animation morphing effect for all digits
+sampler is assumed to be an instance of Sampler()
+'''
+
+def create_z_array():
+  z_array = []
+  for i in range(10):
+    z_array.append(sampler.encode(sampler.get_random_specific_mnist(i)))
+    sampler.show_image_from_z(z_array[i])
+  return z_array
+
+def make_img_data_array(z_array_input, sinusoid = True, fps=24, x_dim = 1080, y_dim = 1080):
+  z_array = list(z_array_input)
+  n = len(z_array)
+  data = []
+  for i in range(n-1):
+    print "Morphing Image #", i
+    sampler.show_image_from_z(z_array[i])
+    data.append(sampler.morph(z_array[i], z_array[i+1], fps, sinusoid = sinusoid, x_dim = x_dim, y_dim = y_dim))
+  sampler.show_image_from_z(z_array[n-1])
+  data.append(sampler.morph(z_array[n-1], z_array[0], fps, sinusoid = sinusoid, x_dim = x_dim, y_dim = y_dim))
+  return data
+
+def write_data_array_as_gif(data_array, filename, fps = 24):
+  result = []
+  for i in range(len(data_array)):
+    result = result + data_array[i]
+  sampler.save_anim_gif(result, filename, 1.0 / float(fps))
+