datumaro.plugins.data_formats.voc.format

Module Attributes

LabelMapConfig

A type representing a label map config

Functions

generate_colormap([length])
make_voc_categories([label_map, task])
make_voc_label_map([task])
parse_label_map(path)	Parses a label map file in the format: 'name : color (r, g, b) : parts (hand, feet, ...) : actions (siting, standing, ...)'
parse_meta_file(path)
write_label_map(path, label_map)
write_meta_file(path, label_map)

Classes

VocAction(value)	An enumeration.
VocBodyPart(value)	An enumeration.
VocImporterType(value)	An enumeration.
VocLabel(value)	An enumeration.
VocPath()
VocPose(value)	An enumeration.
VocTask(value)	An enumeration.

class datumaro.plugins.data_formats.voc.format.VocTask(value)

Bases: Enum

An enumeration.

voc = 1

voc_classification = 2

voc_detection = 3

voc_segmentation = 4

voc_instance_segmentation = 5

voc_action = 6

voc_layout = 7

class datumaro.plugins.data_formats.voc.format.VocLabel(value)

Bases: Enum

An enumeration.

background = 0

aeroplane = 1

bicycle = 2

bird = 3

boat = 4

bottle = 5

bus = 6

car = 7

cat = 8

chair = 9

cow = 10

diningtable = 11

dog = 12

horse = 13

motorbike = 14

person = 15

pottedplant = 16

sheep = 17

sofa = 18

train = 19

tvmonitor = 20

ignored = 255

class datumaro.plugins.data_formats.voc.format.VocPose(value)

Bases: Enum

An enumeration.

Unspecified = 1

Left = 2

Right = 3

Frontal = 4

Rear = 5

class datumaro.plugins.data_formats.voc.format.VocBodyPart(value)

Bases: Enum

An enumeration.

head = 1

hand = 2

foot = 3

class datumaro.plugins.data_formats.voc.format.VocAction(value)

Bases: Enum

An enumeration.

other = 1

jumping = 2

phoning = 3

playinginstrument = 4

reading = 5

ridingbike = 6

ridinghorse = 7

running = 8

takingphoto = 9

usingcomputer = 10

walking = 11

datumaro.plugins.data_formats.voc.format.generate_colormap(length: int = 256) → Dict[int, Tuple[int, int, int]]

class datumaro.plugins.data_formats.voc.format.VocImporterType(value)

Bases: Enum

An enumeration.

default = 1

roboflow = 2

class datumaro.plugins.data_formats.voc.format.VocPath

Bases: object

IMAGES_DIR = 'JPEGImages'

ANNOTATIONS_DIR = 'Annotations'

SEGMENTATION_DIR = 'SegmentationClass'

INSTANCES_DIR = 'SegmentationObject'

SUBSETS_DIR = 'ImageSets'

IMAGE_EXT = '.jpg'

SEGM_EXT = '.png'

LABELMAP_FILE = 'labelmap.txt'

TASK_DIR = {VocTask.voc: 'Main', VocTask.voc_action: 'Action', VocTask.voc_classification: 'Main', VocTask.voc_detection: 'Main', VocTask.voc_instance_segmentation: 'Segmentation', VocTask.voc_layout: 'Layout', VocTask.voc_segmentation: 'Segmentation'}

datumaro.plugins.data_formats.voc.format.LabelMapConfig

A type representing a label map config

alias of Dict[str, Tuple[Optional[Tuple[int, int, int]], List[str], List[str]]]

datumaro.plugins.data_formats.voc.format.make_voc_label_map(task: VocTask | None = None) → Dict[str, Tuple[Tuple[int, int, int] | None, List[str], List[str]]]

datumaro.plugins.data_formats.voc.format.parse_label_map(path: str) → Dict[str, Tuple[Tuple[int, int, int] | None, List[str], List[str]]]

Parses a label map file in the format: ‘name : color (r, g, b) : parts (hand, feet, …) : actions (siting, standing, …)’

Parameters:

path – File path

Returns:

label -> (color, parts, actions)

Return type:

A dictionary

datumaro.plugins.data_formats.voc.format.parse_meta_file(path: str) → Dict[str, Tuple[Tuple[int, int, int] | None, List[str], List[str]]]

datumaro.plugins.data_formats.voc.format.write_label_map(path: str, label_map: Dict[str, Tuple[Tuple[int, int, int] | None, List[str], List[str]]])

datumaro.plugins.data_formats.voc.format.write_meta_file(path: str, label_map: Dict[str, Tuple[Tuple[int, int, int] | None, List[str], List[str]]])

datumaro.plugins.data_formats.voc.format.make_voc_categories(label_map: Dict[str, Tuple[Tuple[int, int, int] | None, List[str], List[str]]] | None = None, task: VocTask | None = VocTask.voc) → Dict[AnnotationType, Categories]