neodroidvision.data.classification.nlet.triplet_dataset.TripletDataset

class neodroidvision.data.classification.nlet.triplet_dataset.TripletDataset(data_path: ~typing.Union[str, ~pathlib.Path], split: ~draugr.numpy_utilities.datasets.splitting.SplitEnum = SplitEnum.training, return_categories: bool = False, *, transform: callable = ToTensor(), target_transform: callable = None, loader=<function default_loader>, is_valid_file: callable = None)

Bases: PairDataset

# This dataset generates a triple of images. an image of a category, another of the same category and lastly one from another category

__init__(data_path: ~typing.Union[str, ~pathlib.Path], split: ~draugr.numpy_utilities.datasets.splitting.SplitEnum = SplitEnum.training, return_categories: bool = False, *, transform: callable = ToTensor(), target_transform: callable = None, loader=<function default_loader>, is_valid_file: callable = None)

Methods

__init__(data_path[, split, ...])
plot_images(images[, label])	param images
sample([horizontal_merge])	description

Attributes

predictor_shape	return: :rtype:
response_shape	return: :rtype:
split_names	return: :rtype:

static plot_images(images, label=None) → None

Parameters

• images –

• label –

property predictor_shape: Tuple[int, ...]

return: :rtype:

property response_shape: Tuple[int, ...]

return: :rtype:

sample(horizontal_merge: bool = False) → None

description

property split_names: Dict[SplitEnum, str]

return: :rtype: