distribution#

Module containing the distribution checks in the vision package.

Classes

TrainTestLabelDrift

Calculate label drift between train dataset and test dataset, using statistical measures.

TrainTestPredictionDrift

Calculate prediction drift between train dataset and test dataset, using statistical measures.

ImageDatasetDrift

Calculate drift between the entire train and test datasets (based on image properties) using a trained model.

HeatmapComparison

Check if the average image brightness (or bbox location if applicable) is similar between train and test set.

ImagePropertyDrift

Calculate drift between train dataset and test dataset per image property, using statistical measures.

ImagePropertyOutliers

Find outliers images with respect to the given properties.

LabelPropertyOutliers

Find outliers labels with respect to the given properties.

NewLabels

Detects labels that apper only in the test set.