Merge remote-tracking branch 'origin/master' into allanzelener/initia…

…l_loss_implementation

README.md

@@ -36,10 +36,6 @@ pip install numpy
 pip install tensorflow-gpu  # CPU-only: conda install -c conda-forge tensorflow
 pip install keras # Possibly older release: conda install keras
 ```
-Until [Keras PR 5350](https://github.com/fchollet/keras/pull/5350) is merged:
- ```
- pip install git+https://github.com/allanzelener/keras.git@func_defaults_to_tuple
- ```
 
 ## Quick Start
 
@@ -71,13 +67,11 @@ YAD2K assumes the Keras backend is Tensorflow. In particular for YOLO_v2 models
 
 ## Known Issues and TODOs
 
-- Error deserializing Lambda wrapping space_to_depth. Apply [this PR to Keras](https://github.com/fchollet/keras/pull/5350).
-  - `pip install git+https://github.com/allanzelener/keras.git@func_defaults_to_tuple`
-- Add YOLO_v2 loss function.
+- Add YOLO_v2 loss function. (In-progress implementation on branch allanzelener/initial_loss_implementation).
 - Script to train YOLO_v2 reference model.
 - Support for additional Darknet layer types.
 - Tuck away the Tensorflow dependencies with Keras wrappers where possible.
-- YOLO_v2 model does not support fully convolutional mode.
+- YOLO_v2 model does not support fully convolutional mode. Current implementation assumes 1:1 aspect ratio images.
 
 ## Darknets of Yore
 

yad2k.py

@@ -19,7 +19,7 @@
 from keras.layers.normalization import BatchNormalization
 from keras.models import Model
 from keras.regularizers import l2
-from keras.utils.visualize_util import plot
+from keras.utils.vis_utils import plot_model as plot
 
 from yad2k.models.keras_yolo import (space_to_depth_x2,
                                      space_to_depth_x2_output_shape)
@@ -238,7 +238,8 @@ def _main(args):
             prev_layer = all_layers[-1]
 
         elif section.startswith('region'):
-            anchors = cfg_parser[section]['anchors']
+            with open('{}_anchors.txt'.format(output_root), 'w') as f:
+                print(cfg_parser[section]['anchors'], file=f)
 
         elif (section.startswith('net') or section.startswith('cost') or
               section.startswith('softmax')):
@@ -261,10 +262,6 @@ def _main(args):
     if remaining_weights > 0:
         print('Warning: {} unused weights'.format(len(remaining_weights)))
 
-    if 'region' in cfg_parser.sections():
-        with open('{}_anchors.txt'.format(output_root), 'w') as f:
-            print(anchors, file=f)
-
     if args.plot_model:
         plot(model, to_file='{}.png'.format(output_root), show_shapes=True)
         print('Saved model plot to {}.png'.format(output_root))