def get_model():
# Real-time data preprocessing
img_prep = ImagePreprocessing()
img_prep.add_samplewise_zero_center()
img_prep.add_samplewise_stdnorm()
net = tflearn.input_data(shape=[None, 400], data_preprocessing=img_prep)
net = tflearn.fully_connected(net, 512, activation='relu')
net = tflearn.fully_connected(net, 512, activation='relu')
net = tflearn.fully_connected(net, 400, activation='sigmoid')
net = tflearn.regression(net, optimizer='adam', loss='binary_crossentropy')
model = tflearn.DNN(net)
return net, model
def createModel(parameters):
img_prep = ImagePreprocessing()
img_prep.add_samplewise_zero_center()
imgaug = tflearn.ImageAugmentation()
imgaug.add_random_flip_leftright()
imgaug.add_random_flip_updown()
imgaug.add_random_90degrees_rotation()
# Building convolutional network
network = input_data(shape=[None, 64, 64, 3], data_preprocessing=img_prep, data_augmentation=imgaug, name='input')
network = conv_2d(network, parameters.CNN1FeatureSize, parameters.CNN1Size, activation='relu', regularizer="L2") # padding same
network=batch_normalization(network)
if parameters.CNN2Size > 0:
network = conv_2d(network, parameters.CNN2FeatureSize, parameters.CNN2Size, activation='relu', regularizer="L2")#2. CNN
network=batch_normalization(network)
if parameters.CNN3Size > 0:
network = conv_2d(network, parameters.CNN3FeatureSize, parameters.CNN3Size, activation='relu', regularizer="L2")#3. CNN
network=batch_normalization(network)
network = max_pool_2d(network, 2)
if parameters.CNN4Size > 0:
network = conv_2d(network, parameters.CNN4FeatureSize, parameters.CNN4Size, activation='relu', regularizer="L2")#4. CNN
network=batch_normalization(network)
network = max_pool_2d(network, 2)
network = conv_2d(network, parameters.CNN4FeatureSize, parameters.CNN4Size, activation='relu', regularizer="L2")#5. CNN
network=batch_normalization(network)
network = max_pool_2d(network, 2)
#if parameters.FullyConn1Size > 0:
# network = fully_connected(network, parameters.FullyConn1Size, activation='relu', regularizer="L2")
# network=batch_normalization(network)
# #network = dropout(network, 0.2)
if parameters.FullyConn2Size > 0:
network = fully_connected(network, parameters.FullyConn1Size, activation='relu', regularizer="L2")
network=batch_normalization(network)
network = dropout(network, 0.2)
if parameters.FullyConn2Size > 0:
network = fully_connected(network, parameters.FullyConn2Size, activation='relu', regularizer="L2")
network=batch_normalization(network)
network = dropout(network, 0.2)
g = tflearn.fully_connected(network, 2, activation='softmax')
g = tflearn.regression(g, optimizer='SGD', loss='categorical_crossentropy', metric='default' , learning_rate=0.01,batch_size=64)
return g
classifier.HEIGHT = HEIGHT
classifier.WIDTH = WIDTH
classifier.IMAGE = HEIGHT
classifier.CHANNELS = CHANNELS
print("\t Path:", args.data_dir)
print("\tShape (train):", X.shape, Y.shape)
print("\tShape (val):", Xv.shape, Yv.shape)
print("Data loaded!\n")
# Real-time data preprocessing (samplewise or featurewise)
img_prep = None
if (args.pproc):
img_prep = ImagePreprocessing()
img_prep.add_samplewise_zero_center()
img_prep.add_samplewise_stdnorm()
#img_prep.add_zca_whitening()
# Real-time data augmentation
img_aug = ImageAugmentation()
img_aug.add_random_flip_leftright()
img_aug.add_random_flip_updown()
img_aug.add_random_rotation(max_angle=5.)
# computational resources definition (made changes on TFLearn's config.py)
tflearn.init_graph(num_cores=8, allow_growth=True)
# network definition
network = input_data(
shape=[None, HEIGHT, WIDTH, CHANNELS], # shape=[None,IMAGE, IMAGE] for RNN
CLASSES = 10
#HEIGHT = 240
#WIDTH = 320
#CHANNELS = 3
# get command line arguments
arch = sys.argv[1] # name of architecture
modelpath = sys.argv[2] # path to saved model
layer = sys.argv[3] # layer name, for example, Conv2D or Conv2D_1
ichannel = int(
sys.argv[4]) # input channel for displaying kernels and convolutions
# Real-time data preprocessing
img_prep = ImagePreprocessing()
img_prep.add_samplewise_zero_center(
) # per sample (featurewise is a global value)
img_prep.add_samplewise_stdnorm() # per sample (featurewise is a global value)
# Real-time data augmentation
img_aug = ImageAugmentation()
img_aug.add_random_flip_leftright()
img_aug.add_random_flip_updown()
img_aug.add_random_rotation(max_angle=10.)
# network definition
network = input_data(
shape=[None, HEIGHT, WIDTH, CHANNELS], # shape=[None,IMAGE, IMAGE] for RNN
data_preprocessing=None,
data_augmentation=None)
#in2 = input_data(shape=[None,1])
