def train(self):
self.X_train, self.X_crop_train, self.X_test, self.X_crop_test = crop_encoder_data.load_data(
0.5)
if K.image_dim_ordering() == 'th':
self.X_train = self.X_train.reshape(self.X_train.shape[0], 1,
self.param_dict['img_rows'],
self.param_dict['img_cols'])
self.X_test = self.X_test.reshape(self.X_test.shape[0], 1,
self.param_dict['img_rows'],
self.param_dict['img_cols'])
self.X_crop_train = self.X_crop_train.reshape(
self.X_train.shape[0], 1, self.param_dict['crop_rows'],
self.param_dict['crop_cols'])
self.X_crop_test = self.X_crop_test.reshape(
self.X_test.shape[0], 1, self.param_dict['crop_rows'],
self.param_dict['crop_cols'])
#input_shape = (1, self.img_rows, img_cols)
else:
self.X_train = self.X_train.reshape(self.X_train.shape[0],
self.param_dict['img_rows'],
self.param_dict['img_cols'], 1)
self.X_test = self.X_test.reshape(self.X_test.shape[0],
self.param_dict['img_rows'],
self.param_dict['img_cols'], 1)
self.X_crop_train = self.X_crop_train.reshape(
self.X_train.shape[0],
self.param_dict['crop_rows'] * self.param_dict['crop_cols'])
self.X_crop_test = X_crop_test.reshape(
X_test.shape[0],
self.param_dict['crop_rows'] * self.param_dict['crop_cols'])
#input_shape = (img_rows,img_cols,1)
self.X_train = self.X_train.astype('float32')
self.X_train /= 255
self.X_test = self.X_test.astype('float32')
self.X_test /= 255
self.X_crop_train = X_crop_train.astype('float32')
self.X_crop_train /= 255
self.X_crop_test = X_crop_test.astype('float32')
self.X_crop_test /= 255
nb_classes = 8
nb_epoch = 10
# input image dimensions
img_rows, img_cols = 128, 128
crop_rows, crop_cols = 32, 32
# number of convolutional filters to use
nb_filters = 16
# size of pooling area for max pooling
pool_size = (2, 2)
# convolution kernel size
kernel_size = (3, 3)
# the data, shuffled and split between train and test sets
X_train, X_crop_train, X_test, X_crop_test = crop_encoder_data.load_data(0.5)
# (X_train, y_train), (X_test, y_test) = mnist.load_data()
if K.image_dim_ordering() == 'th':
X_train = X_train.reshape(X_train.shape[0], 1, img_rows, img_cols)
X_test = X_test.reshape(X_test.shape[0], 1, img_rows, img_cols)
X_crop_train = X_crop_train.reshape(X_train.shape[0], 1, crop_rows,
crop_cols)
X_crop_test = X_crop_test.reshape(X_test.shape[0], 1, crop_rows, crop_cols)
input_shape = (1, img_rows, img_cols)
else:
X_train = X_train.reshape(X_train.shape[0], img_rows, img_cols, 1)
X_test = X_test.reshape(X_test.shape[0], img_rows, img_cols, 1)
X_crop_train = X_crop_train.reshape(X_train.shape[0], crop_rows, crop_cols,
1)
X_crop_test = X_crop_test.reshape(X_test.shape[0], crop_rows, crop_cols, 1)