def get_training_set(data_dir, upscale_factor, patch_size, data_augmentation):
hr_dir = join(data_dir, 'HR')
lr_dir = join(data_dir, 'LR')
return DatasetFromFolder(hr_dir,
lr_dir,
patch_size,
upscale_factor,
data_augmentation,
transform=transform())
def get_test_set(upscale_factor):
root_dir = download_bsd300()
test_dir = join(root_dir, "test")
crop_size = calculate_valid_crop_size(256, upscale_factor)
return DatasetFromFolder(test_dir,
input_transform=input_transform(
crop_size, upscale_factor),
target_transform=target_transform(crop_size))
def create_famous_dataset(test_path, noise, pad):
def pre_process_fn(_x):
return normilize(_x, 255)
def input_process_fn(_x):
return gaussian(_x, is_training=True, mean=0, stddev=normilize(noise, 255))
return DatasetFromFolder(
test_path,
pre_transform=pre_process_fn,
use_cuda=USE_CUDA,
inputs_transform=input_process_fn
)
def create_test_dataset(test_path, noise, pad):
def pre_process_fn(_x):
return normilize(_x, 255)
def input_process_fn(_x):
return gaussian(_x, is_training=True, mean=0, stddev=normilize(noise, 255))
file_of_filenames =\
os.path.join(common.project_dir(), 'pascal2010_test_imgs.txt')
return DatasetFromFolder(
test_path,
file_of_filenames=file_of_filenames,
pre_transform=pre_process_fn,
use_cuda=USE_CUDA,
inputs_transform=input_process_fn
)