object detection

We consider the object detection AI problem to encompass models to infer bounding boxes, labels (from a predefined label set), and scores for objects in an image.

primitives

object_detection.InputType	Role alias for model/dataset input in the object-detection protocol family.
object_detection.TargetType	Role alias for model/dataset target in the object-detection protocol family.
object_detection.DatumMetadataType	Role alias for datum-level metadata in object-detection protocol signatures.
object_detection.Image	Semantic alias for a single image datum.
object_detection.ObjectDetectionTarget(...)	An object-detection target protocol.
object_detection.DatumMetadata	Semantic alias for per-datum metadata in object-detection tasks.

components

object_detection.Augmentation(*args, **kwargs)	An augmentation protocol for the object detection AI problem.
object_detection.DataLoader(*args, **kwargs)	A dataloader protocol for the object detection AI problem providing batch-level data access.
object_detection.Dataset(*args, **kwargs)	A dataset protocol for object detection AI problem providing datum-level data access.
object_detection.FieldwiseDataset(*args, ...)	A specialization of Dataset protocol (i.e., a subprotocol) that specifies additional accessor methods for getting input, target, and metadata individually.
object_detection.Metric(*args, **kwargs)	A metric protocol for the object detection AI problem.
object_detection.Model(*args, **kwargs)	A model protocol for the object detection AI problem.