def __init__(
self,
source_root,
target_root,
opts,
latents_root=None,
labels_path=None,
target_transform=None,
source_transform=None,
):
self.target_paths = sorted(data_utils.make_dataset(target_root))
self.source_paths = sorted(data_utils.make_dataset(source_root))
self.latent_paths = None
if latents_root is not None:
self.latent_paths = sorted(
data_utils.make_latents_dataset(latents_root)
)
self.path_to_label = None
if labels_path is not None:
with open(labels_path) as f:
labels = json.load(f)["labels"]
self.path_to_label = {path: label for path, label in labels}
self.source_transform = source_transform
self.target_transform = target_transform
self.opts = opts
def __init__(self,
source_root,
target_root,
opts,
target_transform=None,
source_transform=None):
self.source_paths = sorted(data_utils.make_dataset(source_root))
self.target_paths = sorted(data_utils.make_dataset(target_root))
self.source_transform = source_transform
self.target_transform = target_transform
self.opts = opts
def __init__(self, root=None, paths_list=None, opts=None, transform=None, return_path=False): if paths_list is None: self.paths = sorted(data_utils.make_dataset(root)) else: self.paths = data_utils.make_dataset_from_paths_list(paths_list) self.transform = transform self.opts = opts self.return_path = return_path
# ===========================
if __name__ == '__main__':
args = argument_parser()
input_path = args.input_path
output_path = args.output_path
if input_path == 'voc_horse':
# get all non-horse voc files:
input_filenames = get_voc_classification_filenames(voc_folder_path=CONSTS.VOC_DIR, category='horse')
elif input_path == 'voc_mis':
# get all non-person voc files:
input_filenames = get_voc_classification_filenames(voc_folder_path=CONSTS.VOC_DIR, category='person')
else:
input_filenames = make_dataset(dir=input_path, ext='jpg')
gen_data(mode=args.mode, input_filenames=input_filenames, output_path=output_path,
num_sets=args.num_sets, minimalImage_size=args.minimalImage_size, limit=args.limit)
# [Filter box proposals with the selective search code]
# # Feel free to change parameters
# boxes_filter = selective_search.box_filter(boxes, min_size=20, topN=80)
#
# # draw rectangles on the original image
# fig, ax = plt.subplots(figsize=(6, 6))
# ax.imshow(image)
# for x1, y1, x2, y2 in boxes_filter:
def __init__(self, root, opts, transform=None): self.paths = sorted(data_utils.make_dataset(root)) self.transform = transform self.opts = opts
