def CCPDGreen(path: str) -> Dataset:
"""`CCPDGreen <https://github.com/detectRecog/CCPD>`_ dataset.
The file structure should be like::
<path>
ccpd_green/
train/
test/
val/
Arguments:
path: The root directory of the dataset.
Returns:
Loaded :class: `~tensorbay.dataset.dataset.Dataset` instance.
"""
root_path = os.path.join(os.path.abspath(os.path.expanduser(path)), "ccpd_green")
dataset = Dataset(DATASET_NAME_CCPDGREEN)
dataset.load_catalog(os.path.join(os.path.dirname(__file__), "catalog.json"))
for segment_name in _CCPDGREEN_SEGMENTS:
segment = dataset.create_segment(segment_name)
for image_path in glob(os.path.join(root_path, segment_name, "*.jpg")):
data = Data(image_path)
data.label.polygon = _get_polygons(image_path)
segment.append(data)
return dataset
def _get_ccpd_image_path(root_path: str, segment_head: str, segment_tail: str) -> Iterator[str]:
if segment_tail == "base":
file_path = os.path.join(root_path, "splits", f"{segment_head}.txt")
with open(file_path, encoding="utf-8") as fp:
for image_path in fp:
yield os.path.join(root_path, image_path.strip())
else:
for image_path in glob(os.path.join(root_path, f"ccpd_{segment_tail}", "*.jpg")):
yield image_path
def OxfordIIITPet(path: str) -> Dataset:
"""`OxfordIIITPet <https://www.robots.ox.ac.uk/~vgg/data/pets/>`_ dataset.
The file structure should be like::
<path>
annotations/
trimaps/
Bombay_113.png
Bombay_114.png
...
xmls/
Birman_174.xml
Birman_175.xml
...
list.txt
test.txt
trainval.txt
README
images/
Bombay_117.jpg
Bombay_118.jpg
...
Arguments:
path: The root directory of the dataset.
Returns:
Loaded :class: `~tensorbay.dataset.dataset.Dataset` instance.
"""
root_path = os.path.abspath(os.path.expanduser(path))
dataset = Dataset(DATASET_NAME)
dataset.load_catalog(
os.path.join(os.path.dirname(__file__), "catalog.json"))
trainval_segment = dataset.create_segment("trainval")
test_segment = dataset.create_segment("test")
annotation_path = os.path.join(root_path, "annotations")
for image_path in glob(os.path.join(root_path, "images", "*.jpg")):
image_name = os.path.splitext(os.path.basename(image_path))[0]
name = "Cat" if image_name.istitle() else "Dog"
category, num = image_name.rsplit("_", 1)
data = Data(image_path,
target_remote_path=f"{category}_{num.zfill(3)}.jpg")
label = data.label
label.classification = Classification(category=f"{name}.{category}")
label.semantic_mask = SemanticMask(
os.path.join(annotation_path, "trimaps", f"{image_name}.png"))
xml_path = os.path.join(annotation_path, "xmls", f"{image_name}.xml")
if os.path.exists(xml_path):
label.box2d = _get_box_label(xml_path)
trainval_segment.append(data)
else:
test_segment.append(data)
return dataset
def COCO2017(path: str) -> Dataset:
"""`COCO2017 <https://cocodataset.org/#home>`_ dataset.
The file structure should be like::
<path>
annotations/
panoptic_train2017/
000000116037.png
000000116040.png
...
panoptic_val2017/
instances_train2017.json
instances_val2017.json
panoptic_train2017.json
panoptic_val2017.json
person_keypoints_train2017.json
person_keypoints_val2017.json
train2017/
000000116026.jpg
000000116031.jpg
...
test2017/
val2017/
unlabeled2017/
Arguments:
path: The root directory of the dataset.
Returns:
Loaded :class: `~tensorbay.dataset.dataset.Dataset` instance.
"""
root_path = os.path.abspath(os.path.expanduser(path))
dataset = Dataset(DATASET_NAME)
dataset.load_catalog(
os.path.join(os.path.dirname(__file__), "catalog.json"))
for segment_name in _LABELED_SEGMENT_NAMES:
segment = dataset.create_segment(segment_name)
annotation_path = os.path.join(root_path, "annotations")
task_information = _get_information(annotation_path, segment_name)
categories = task_information["categories"]
for image_path in glob(
os.path.join(root_path, f"{segment_name}2017", "*.jpg")):
data = Data(image_path)
image_stem = os.path.splitext(os.path.basename(image_path))[0]
image_id = int(image_stem)
label = _get_instance_label(
task_information["instances_annotations"], image_id,
categories)
label.keypoints2d = _get_keypoints2d(
task_information["person_keypoints_annotations"], image_id,
categories)
label.box2d.extend(
_get_panoptic_box2d(task_information["panoptic_annotations"],
image_id, categories))
label.panoptic_mask = _get_panoptic_mask(
annotation_path,
segment_name,
image_stem,
task_information["panoptic_annotations"],
image_id,
)
data.label = label
segment.append(data)
for segment_name in _UNLABELED_SEGMENT_NAMES:
segment = dataset.create_segment(segment_name)
for image_path in glob(
os.path.join(root_path, f"{segment_name}2017", "*.jpg")):
segment.append(Data(image_path))
return dataset