def __init__(self, model, fn_A, fn_B, batch_size=64):
self.batch_size = batch_size
self.model = model
generator = TrainingDataGenerator(self.random_transform_args, 160)
self.images_A = generator.minibatchAB(fn_A, self.batch_size)
self.images_B = generator.minibatchAB(fn_B, self.batch_size)
def __init__(self, model, fn_A, fn_B, batch_size, *args):
self.batch_size = batch_size
self.model = model
generator = TrainingDataGenerator(self.random_transform_args, 160)
self.images_A = generator.minibatchAB(fn_A, self.batch_size)
self.images_B = generator.minibatchAB(fn_B, self.batch_size)
def __init__(self, model, fn_A, fn_B, batch_size, *args):
self.batch_size = batch_size
self.model = model
from timeit import default_timer as clock
self._clock = clock
generator = TrainingDataGenerator(self.random_transform_args, 160, 5, zoom=self.model.IMAGE_SHAPE[0]//64)
self.images_A = generator.minibatchAB(fn_A, self.batch_size)
self.images_B = generator.minibatchAB(fn_B, self.batch_size)
self.generator = generator
def __init__(self, model, fn_A, fn_B, batch_size, *args):
self.batch_size = batch_size
self.model = model
from timeit import default_timer as clock
self._clock = clock
generator = TrainingDataGenerator(self.random_transform_args, 160, 5, zoom=2)
self.images_A = generator.minibatchAB(fn_A, self.batch_size)
self.images_B = generator.minibatchAB(fn_B, self.batch_size)
self.generator = generator
def load_generator(self):
""" Pass arguments to TrainingDataGenerator and return object """
input_size = self.model.input_shape[0]
output_size = self.model.output_shape[0]
generator = TrainingDataGenerator(input_size, output_size,
self.model.training_opts)
return generator
def load_generator(self):
""" Pass arguments to TrainingDataGenerator and return object """
logger.debug("Loading generator: %s", self.side)
input_size = self.model.input_shape[0]
output_size = self.model.output_shape[0]
logger.debug("input_size: %s, output_size: %s", input_size, output_size)
generator = TrainingDataGenerator(input_size, output_size, self.model.training_opts)
return generator
def __init__(self, model, fn_A, fn_B, batch_size, *args):
self.batch_size = batch_size
self.model = model
#generator = TrainingDataGenerator(self.random_transform_args, 160)
# make sre to keep zoom=2 or you won't get 128x128 vectors as input
#generator = TrainingDataGenerator(self.random_transform_args, 220, 5, zoom=2)
generator = TrainingDataGenerator(self.random_transform_args,
160,
6,
zoom=2)
#generator = TrainingDataGenerator(self.random_transform_args, 180, 7, zoom=2)
self.images_A = generator.minibatchAB(fn_A, self.batch_size)
self.images_B = generator.minibatchAB(fn_B, self.batch_size)
self.generator = generator
def get_image_data(self, input_images, batch_size):
""" Get training images """
random_transform_args = {
'rotation_range': 0,
'zoom_range': 0,
'shift_range': 0,
'random_flip': 0
}
zoom = 1
if hasattr(self.trainer.model, 'IMAGE_SHAPE'):
zoom = self.trainer.model.IMAGE_SHAPE[0] // 64
generator = TrainingDataGenerator(random_transform_args, 160, zoom)
batch = generator.minibatchAB(input_images, batch_size,
doShuffle=False)
return next(batch)[2]
def _load_generator(self):
input_size = self._model.model.input_shape[1] # model.input_shape = (None, n)
generator = TrainingDataGenerator(input_size)
return generator
