The function WolfDetectorTrainingDataset currently packs box information into flat arrays that are used as 'labels' for training, returning a 'trainingDataset'.
This would be a perfect place to apply image transformations to the images and boxes, color shifts, interesting color effects, etc., thereby turning a small dataset into a larger dataset.
The standard library of Wolfram Language already has support to perform geometric transformations and image effects, so it is only a matter of adding it to data data processing utilities.
The function
WolfDetectorTrainingDataset
currently packs box information into flat arrays that are used as 'labels' for training, returning a 'trainingDataset'.This would be a perfect place to apply image transformations to the images and boxes, color shifts, interesting color effects, etc., thereby turning a small dataset into a larger dataset.
The standard library of Wolfram Language already has support to perform geometric transformations and image effects, so it is only a matter of adding it to data data processing utilities.