updatae

Author: foamliu

config.py

@@ -1,4 +1,4 @@
-img_width, img_height = 640, 640
+img_width, img_height = 224, 224
 num_channels = 3
 train_data = 'data/train'
 valid_data = 'data/valid'

demo.py

@@ -11,7 +11,7 @@
 from utils import draw_str
 
 if __name__ == '__main__':
-    img_width, img_height = 640, 640
+    img_width, img_height = 224, 224
     num_channels = 3
     num_classes = 80
 

resnet_50.py

@@ -1,11 +1,10 @@
 # -*- coding: utf-8 -*-
 
-from keras.optimizers import SGD
-from keras.layers import Input, Dense, Conv2D, MaxPooling2D, AveragePooling2D, ZeroPadding2D, Flatten, Activation, add
+from keras import backend as K
+from keras.layers import Input, Dense, Conv2D, MaxPooling2D, AveragePooling2D, ZeroPadding2D, Flatten, Activation, add, \
+    Dropout
 from keras.layers.normalization import BatchNormalization
 from keras.models import Model
-from keras import backend as K
-
 from sklearn.metrics import log_loss
 
 
@@ -80,7 +79,7 @@ def conv_block(input_tensor, kernel_size, filters, stage, block, strides=(2, 2))
     return x
 
 
-def resnet50_model_old(img_rows, img_cols, color_type=1, num_classes=None):
+def resnet50_model(img_rows, img_cols, color_type=1, num_classes=None):
     """
     Resnet 50 Model for Keras
 
@@ -149,71 +148,15 @@ def resnet50_model_old(img_rows, img_cols, color_type=1, num_classes=None):
 
     model.load_weights(weights_path)
 
-    return model
-
-
-def resnet50_model_new(img_rows, img_cols, color_type, num_classes):
-    """
-    Resnet 50 Model for Keras
-
-    Model Schema is based on
-    https://github.com/fchollet/deep-learning-models/blob/master/resnet50.py
-
-    ImageNet Pretrained Weights
-    https://github.com/fchollet/deep-learning-models/releases/download/v0.2/resnet50_weights_th_dim_ordering_th_kernels.h5
-
-    Parameters:
-      img_rows, img_cols - resolution of inputs
-      channel - 1 for grayscale, 3 for color
-      num_classes - number of class labels for our classification task
-    """
-
-    # Handle Dimension Ordering for different backends
-    global bn_axis
-    if K.image_dim_ordering() == 'tf':
-        bn_axis = 3
-        img_input = Input(shape=(img_rows, img_cols, color_type))
-    else:
-        bn_axis = 1
-        img_input = Input(shape=(color_type, img_rows, img_cols))
-
-    x = ZeroPadding2D((3, 3))(img_input)
-    x = Conv2D(64, (7, 7), strides=(2, 2), name='conv1')(x)
-    x = BatchNormalization(axis=bn_axis, name='bn_conv1')(x)
-    x = Activation('relu')(x)
-    x = MaxPooling2D((3, 3), strides=(2, 2))(x)
-
-    x = conv_block(x, 3, [64, 64, 256], stage=2, block='a', strides=(1, 1))
-    x = identity_block(x, 3, [64, 64, 256], stage=2, block='b')
-    x = identity_block(x, 3, [64, 64, 256], stage=2, block='c')
-
-    x = conv_block(x, 3, [128, 128, 512], stage=3, block='a')
-    x = identity_block(x, 3, [128, 128, 512], stage=3, block='b')
-    x = identity_block(x, 3, [128, 128, 512], stage=3, block='c')
-    x = identity_block(x, 3, [128, 128, 512], stage=3, block='d')
-
-    x = conv_block(x, 3, [256, 256, 1024], stage=4, block='a')
-    x = identity_block(x, 3, [256, 256, 1024], stage=4, block='b')
-    x = identity_block(x, 3, [256, 256, 1024], stage=4, block='c')
-    x = identity_block(x, 3, [256, 256, 1024], stage=4, block='d')
-    x = identity_block(x, 3, [256, 256, 1024], stage=4, block='e')
-    x = identity_block(x, 3, [256, 256, 1024], stage=4, block='f')
-
-    x = conv_block(x, 3, [512, 512, 2048], stage=5, block='a')
-    x = identity_block(x, 3, [512, 512, 2048], stage=5, block='b')
-    x = identity_block(x, 3, [512, 512, 2048], stage=5, block='c')
-
+    # Truncate and replace softmax layer for transfer learning
+    # Cannot use model.layers.pop() since model is not of Sequential() type
+    # The method below works since pre-trained weights are stored in layers but not in the model
     x_newfc = AveragePooling2D((7, 7), name='avg_pool')(x)
     x_newfc = Flatten()(x_newfc)
-    x_newfc = Dense(num_classes, activation='softmax', name='fc10')(x_newfc)
+    x_newfc = Dense(num_classes, activation='softmax', name='fc8')(x_newfc)
 
-    # Create another model with our customized softmax
     model = Model(img_input, x_newfc)
 
-    # Learning rate is changed to 0.001
-    sgd = SGD(lr=1e-3, decay=1e-6, momentum=0.9, nesterov=True)
-    model.compile(optimizer=sgd, loss='categorical_crossentropy', metrics=['accuracy'])
-
     return model
 
 

train.py

@@ -1,4 +1,5 @@
 import argparse
+
 import keras
 import tensorflow as tf
 from keras.callbacks import ModelCheckpoint, EarlyStopping
@@ -9,8 +10,7 @@
 
 from config import img_height, img_width, batch_size, patience, num_channels, num_classes, train_data, valid_data, \
     num_train_samples, num_valid_samples, num_epochs, verbose
-from migrate import migrate_model
-from resnet_50 import resnet50_model_new
+from resnet_50 import resnet50_model
 from utils import get_available_gpus, get_available_cpus
 
 if __name__ == '__main__':
@@ -58,27 +58,19 @@ def on_epoch_end(self, epoch, logs=None):
     num_gpu = len(get_available_gpus())
     if num_gpu >= 2:
         with tf.device("/cpu:0"):
+            model = resnet50_model(img_rows=img_height, img_cols=img_width, color_type=num_channels,
+                                   num_classes=num_classes)
             if pretrained_path is not None:
-                model = resnet50_model_new(img_rows=img_height, img_cols=img_width, color_type=num_channels,
-                                           num_classes=num_classes)
                 model.load_weights(pretrained_path)
-            else:
-                model = resnet50_model_new(img_rows=img_height, img_cols=img_width, color_type=num_channels,
-                                           num_classes=num_classes)
-                migrate_model(model)
 
         new_model = multi_gpu_model(model, gpus=num_gpu)
         # rewrite the callback: saving through the original model and not the multi-gpu model.
         model_checkpoint = MyCbk(model)
     else:
+        new_model = resnet50_model(img_rows=img_height, img_cols=img_width, color_type=num_channels,
+                                   num_classes=num_classes)
         if pretrained_path is not None:
-            new_model = resnet50_model_new(img_rows=img_height, img_cols=img_width, color_type=num_channels,
-                                       num_classes=num_classes)
             new_model.load_weights(pretrained_path)
-        else:
-            new_model = resnet50_model_new(img_rows=img_height, img_cols=img_width, color_type=num_channels,
-                                       num_classes=num_classes)
-            migrate_model(new_model)
 
     sgd = SGD(lr=1e-3, decay=1e-6, momentum=0.9, nesterov=True)
     new_model.compile(optimizer=sgd, loss='categorical_crossentropy', metrics=['accuracy'])