dataeval.utils.data¶

Provides utility functions for interacting with Computer Vision datasets.

Submodules¶

`datasets`	Provides access to common Computer Vision datasets.

`Embeddings`	Collection of image embeddings from a dataset.
`Images`	Collection of image data from a dataset.
`Metadata`	Class containing binned metadata.
`Targets`	Dataclass defining targets for image classification or object detection.

`SplitDatasetOutput`	Output class containing test indices and a list of TrainValSplits.

`split_dataset`(labels[, num_folds, stratify, split_on, ...])	Top level splitting function. Returns a dataclass containing a list of train and validation indices.