update gan notes

README.md

@@ -60,7 +60,7 @@ The components are designed to be independent. You can use Model or DataFlow in
 pip install --user -r requirements.txt
 pip install --user -r opt-requirements.txt (some optional dependencies, you can install later if needed)
 ```
-+ Enable `import tensorpack`:
++ Enable `import tensorpack` (or use `greadlink` from `coreutils` brew package if you're on OSX):
 ```
 export PYTHONPATH=$PYTHONPATH:`readlink -f path/to/tensorpack`
 ```

examples/GAN/GAN.py

@@ -42,6 +42,11 @@ class RandomZData(DataFlow):
             yield [np.random.uniform(-1, 1, size=self.shape)]
 
 def build_GAN_losses(vecpos, vecneg):
+    """
+    :param vecpos, vecneg: output of the discriminator (logits) for real
+        and fake images.
+    :return: (loss of G, loss of D)
+    """
     sigmpos = tf.sigmoid(vecpos)
     sigmneg = tf.sigmoid(vecneg)
     tf.summary.histogram('sigmoid-pos', sigmpos)

examples/GAN/README.md

@@ -8,7 +8,7 @@ Reproduce the following GAN-related papers:
 
 + InfoGAN: Interpretable Representation Learning by Information Maximizing GAN. [paper](https://arxiv.org/abs/1606.03657)
 
-See the docstring in each script for detailed usage.
+Detailed usage is in the docstring of each script.
 
 ## DCGAN-CelebA.py
 
@@ -26,12 +26,16 @@ Play with the [pretrained model](https://drive.google.com/drive/folders/0B9IPQTv
 
 ## Image2Image.py
 
-Reproduce Image-to-Image following the setup in [pix2pix](https://github.com/phillipi/pix2pix).
+Image-to-Image following the setup in [pix2pix](https://github.com/phillipi/pix2pix).
 It requires the datasets released by the original authors.
 
+With the cityscapes dataset, it learns to generate semantic segmentation map of urban scene:
+
+![im2im](demo/im2im-cityscapes.jpg)
+
 ## InfoGAN-mnist.py
 
-Reproduce a mnist experiement in InfoGAN.
+Reproduce one mnist experiement in InfoGAN.
 By assuming 10 latent variables corresponding to a categorical distribution and maximizing mutual information,
 the network learns to map the 10 variables to 10 digits in a completely unsupervised way.