Support uint16 image type for ImageFolder - GitHub
github.com › tensorflow › datasetsSep 23, 2020 · The ImageFolder loads the images from a given directory default in uint8. For various reasons it would be useful to have the ability to select the default image dtype ( uint8 or uint16) of the ImageFolder, especially because tfds.features.Image has a uint16 support. Link to the relevant code blocks: split_dict = split_lib.
ImageFolder — Torchvision main documentation
pytorch.org › torchvisionImageFolder. A generic data loader where the images are arranged in this way by default: This class inherits from DatasetFolder so the same methods can be overridden to customize the dataset. root ( string) – Root directory path. transform ( callable, optional) – A function/transform that takes in an PIL image and returns a transformed version.