If using Windows, we recommend using 7-Zip to unzip the exported labels.
All coordinates in the COCO format correspond directly to pixel values. Refer to the COCO documentation for details on the COCO format.
The export contains a list of JSON objects, with each JSON object representing a single data unit and containing detailed information for all exported labels. The example below describes all the keys for a single exported data unit.
KeyDescriptionNotes
annotationsContains all label information
areaThe area contained by the labelin units of square pixels
attributesA list of attributes belonging to a label
classificationsAll radio and checklist attributes and their answers
#relationThe hash of the ontology object the label is related toThe hash is the object’s unique identifier
encord_track_uuidA unique identifier for the object instanceThe same as the objectHash in a JSON export.
manual_annotationIndicates whether the annotation was created manually.True is the annotation is manual, False if created using automated labeling techniques
track_idThe unique identifier for the instance the label is associated withThe value is the same for all annotations belonging to the same instance
bboxPixel coordinates for the top-left corner of a bounding box that encloses the label. This is provided for all shape types.
category_idA numerical ID linking this annotation to a field in the categories list containing Ontology information.For example, the category_id value 1 in an annotation links it to the corresponding Ontology shape with the ID 1 in the categories field of the export.
idThe unique identifier for the labelEach label has a unique identifier, irrespective of which instance it is associated with
image_idThe ID of the image or frame the label is onFor example, ID 0 indicates that the label is on the first image in the images list of the export
iscrowdThe iscrowd field specifies whether the annotation represents a single object (iscrowd=0, using polygons) or a collection of objects (iscrowd=1, using RLE). Single objects may require multiple polygons if occluded, while crowd annotations are used for large groups of objects (e.g., a crowd of people)
keypointsThe keypoints field is a list of coordinates and visibility flags for specific points on an object
num_keypointsThe num_keypoints field indicates the number of labeled keypoints (v > 0) for an object. Objects without labeled keypoints (e.g., small or crowd objects) have num_keypoints=0Keypoints and their connections (skeletons) are defined in the categories section, which includes the names and connectivity of the keypoints for visualization
segmentationContains the (x,y) pixel coordinates for all label vertices
categoriesContains Ontology information
nameThe object’s name in the Ontology
supercategoryThe name of the object’s class
keypointskeypoint for keypoint labels. Omitted for all other labels
skeletonDefines connectivity via a list of keypoint edge pairs and is used for visualization
imagesContains file information for all modalities
coco_urlURL to the file if it is hosted on the COCO dataset’s official serversOnly meaningful when working with the official COCO tasks. In all other cases the coco_url field is empty.
file_nameThe name of the file
heightThe height of the image or frameMeasured in pixels
image_titleThe title of the file
widthThe width of the image or frameMeasured in pixels
infoContains additional information