Module datatap.template.instance_template

Classes

class InstanceTemplate (*, id: bool = False, bounding_box: bool = False, segmentation: bool = False, keypoints: AbstractSet[str] = set(), attributes: Mapping[str, AbstractSet[str]] = {})

Describes how an individual instance is structured.

Expand source code

class InstanceTemplate():
        """
        Describes how an individual instance is structured.
        """

        id: bool
        """
        If `id` is `True`, then all corresponding `Instance`s will have an ID
        that uniquely identifies the object represented by the instance in the
        context of the containing annotation.
        """

        bounding_box: bool
        """
        If `bounding_box` is `True`, then all corresponding `Instance`s will have a
        `BoundingBox` representing the bounds of their shape.
        """

        segmentation: bool
        """
        If `segmentation` is `True`, then all corresponding `Instance`s will have a
        `Segmentation` tightly representing their shape.
        """

        keypoints: AbstractSet[str]
        """
        For each keypoint name specified in `keypoints`, all corresponding instances
        will have a corresponding key in their `keypoints` field, the value of which
        will contain he keypoint if it is present or has an inferrable position in
        the image or `None` if it is not in-frame.
        """

        attributes: Mapping[str, AbstractSet[str]]
        """
        For each attribute name specified in `attributes`, all corresponding
        `Instance`s will provide one of the given values.
        """

        def __init__(
                self,
                *,
                id: bool = False,
                bounding_box: bool = False,
                segmentation: bool = False,
                keypoints: AbstractSet[str] = set(),
                attributes: Mapping[str, AbstractSet[str]] = dict(),
        ):
                self.id = id
                self.bounding_box = bounding_box
                self.segmentation = segmentation
                self.keypoints = keypoints
                self.attributes = attributes

        def to_json(self) -> InstanceTemplateJson:
                """
                Serializes this object as JSON.
                """
                json = InstanceTemplateJson()

                if self.id: json["id"] = True
                if self.bounding_box: json["boundingBox"] = True
                if self.segmentation: json["segmentation"] = True
                if len(self.keypoints) > 0: json["keypoints"] = list(self.keypoints)
                if len(self.attributes) > 0: json["attributes"] = { key: list(values) for key, values in self.attributes.items() }

                return json

        @staticmethod
        def from_json(json: InstanceTemplateJson) -> InstanceTemplate:
                """
                Deserializes a JSON object as an `InstanceTemplate`.
                """
                id = json.get("id", False)
                bounding_box = json.get("boundingBox", False)
                segmentation = json.get("segmentation", False)
                keypoints = set(json.get("keypoints", []))
                attributes = {
                        key: set(values)
                        for key, values in json.get("attributes", {}).items()
                }
                return InstanceTemplate(
                        id = id,
                        bounding_box=bounding_box,
                        segmentation=segmentation,
                        keypoints=keypoints,
                        attributes=attributes,
                )

        def __repr__(self) -> str:
                return basic_repr(
                        "InstanceTemplate",
                        id = self.id,
                        bounding_box = self.bounding_box,
                        segmentation = self.segmentation,
                        keypoints = self.keypoints,
                        attributes = self.attributes,
                )

Class variables

var attributes : Mapping[str, AbstractSet[str]]

For each attribute name specified in attributes, all corresponding Instances will provide one of the given values.

var bounding_box : bool

If bounding_box is True, then all corresponding Instances will have a BoundingBox representing the bounds of their shape.

var id : bool

If id is True, then all corresponding Instances will have an ID that uniquely identifies the object represented by the instance in the context of the containing annotation.

var keypoints : AbstractSet[str]

For each keypoint name specified in keypoints, all corresponding instances will have a corresponding key in their keypoints field, the value of which will contain he keypoint if it is present or has an inferrable position in the image or None if it is not in-frame.

var segmentation : bool

If segmentation is True, then all corresponding Instances will have a Segmentation tightly representing their shape.

Static methods

def from_json(json: InstanceTemplateJson) ‑> InstanceTemplate

Deserializes a JSON object as an InstanceTemplate.

Expand source code

@staticmethod
def from_json(json: InstanceTemplateJson) -> InstanceTemplate:
        """
        Deserializes a JSON object as an `InstanceTemplate`.
        """
        id = json.get("id", False)
        bounding_box = json.get("boundingBox", False)
        segmentation = json.get("segmentation", False)
        keypoints = set(json.get("keypoints", []))
        attributes = {
                key: set(values)
                for key, values in json.get("attributes", {}).items()
        }
        return InstanceTemplate(
                id = id,
                bounding_box=bounding_box,
                segmentation=segmentation,
                keypoints=keypoints,
                attributes=attributes,
        )

Methods

def to_json(self) ‑> InstanceTemplateJson

Serializes this object as JSON.

Expand source code

def to_json(self) -> InstanceTemplateJson:
        """
        Serializes this object as JSON.
        """
        json = InstanceTemplateJson()

        if self.id: json["id"] = True
        if self.bounding_box: json["boundingBox"] = True
        if self.segmentation: json["segmentation"] = True
        if len(self.keypoints) > 0: json["keypoints"] = list(self.keypoints)
        if len(self.attributes) > 0: json["attributes"] = { key: list(values) for key, values in self.attributes.items() }

        return json

class InstanceTemplateJson (*args, **kwargs)

The serialized JSON representation of an instance template.

Expand source code

class InstanceTemplateJson(TypedDict, total=False):
        """
        The serialized JSON representation of an instance template.
        """

        id: bool
        boundingBox: bool
        segmentation: bool
        keypoints: List[str]
        attributes: Dict[str, List[str]]

Ancestors

• builtins.dict

Class variables

var attributes : Dict[str, List[str]]

var boundingBox : bool

var id : bool

var keypoints : List[str]

var segmentation : bool