anomalib.data.folder

Custom Folder Dataset.

This script creates a custom dataset from a folder.

Module Contents

Classes

FolderDataset

Folder Dataset.

Folder

Folder Lightning Data Module.

Functions

_check_and_convert_path(path: Union[str, pathlib.Path]) → pathlib.Path

Check an input path, and convert to Pathlib object.

_prepare_files_labels(path: Union[str, pathlib.Path], path_type: str, extensions: Optional[Tuple[str, Ellipsis]] = None) → Tuple[list, list]

Return a list of filenames and list corresponding labels.

make_dataset(normal_dir: Union[str, pathlib.Path], abnormal_dir: Union[str, pathlib.Path], normal_test_dir: Optional[Union[str, pathlib.Path]] = None, mask_dir: Optional[Union[str, pathlib.Path]] = None, split: Optional[str] = None, split_ratio: float = 0.2, seed: int = 0, create_validation_set: bool = True, extensions: Optional[Tuple[str, Ellipsis]] = None)

Make Folder Dataset.

Attributes

logger

anomalib.data.folder.logger[source]
anomalib.data.folder._check_and_convert_path(path: Union[str, pathlib.Path]) → pathlib.Path[source]

Check an input path, and convert to Pathlib object.

Parameters

path (Union[str, Path]) – Input path.

Returns

Output path converted to pathlib object.

Return type

Path

anomalib.data.folder._prepare_files_labels(path: Union[str, pathlib.Path], path_type: str, extensions: Optional[Tuple[str, Ellipsis]] = None) → Tuple[list, list][source]

Return a list of filenames and list corresponding labels.

Parameters

  • path (Union[str, Path]) – Path to the directory containing images.

  • path_type (str) – Type of images in the provided path (“normal”, “abnormal”, “normal_test”)

  • extensions (Optional[Tuple[str, ...]], optional) – Type of the image extensions to read from the directory.

Returns

Filenames of the images provided in the paths, labels of the images provided in the paths

Return type

List, List

anomalib.data.folder.make_dataset(normal_dir: Union[str, pathlib.Path], abnormal_dir: Union[str, pathlib.Path], normal_test_dir: Optional[Union[str, pathlib.Path]] = None, mask_dir: Optional[Union[str, pathlib.Path]] = None, split: Optional[str] = None, split_ratio: float = 0.2, seed: int = 0, create_validation_set: bool = True, extensions: Optional[Tuple[str, Ellipsis]] = None)[source]

Make Folder Dataset.

Parameters

  • normal_dir (Union[str, Path]) – Path to the directory containing normal images.

  • abnormal_dir (Union[str, Path]) – Path to the directory containing abnormal images.

  • normal_test_dir (Optional[Union[str, Path]], optional) – Path to the directory containing normal images for the test dataset. Normal test images will be a split of normal_dir if None. Defaults to None.

  • mask_dir (Optional[Union[str, Path]], optional) – Path to the directory containing the mask annotations. Defaults to None.

  • split (Optional[str], optional) – Dataset split (ie., either train or test). Defaults to None.

  • split_ratio (float, optional) – Ratio to split normal training images and add to the test set in case test set doesn’t contain any normal images. Defaults to 0.2.

  • seed (int, optional) – Random seed to ensure reproducibility when splitting. Defaults to 0.

  • create_validation_set (bool, optional) – Boolean to create a validation set from the test set. Those wanting to create a validation set could set this flag to True.

  • extensions (Optional[Tuple[str, ...]], optional) – Type of the image extensions to read from the directory.

Returns

an output dataframe containing samples for the requested split (ie., train or test)

Return type

DataFrame

class anomalib.data.folder.FolderDataset(normal_dir: Union[pathlib.Path, str], abnormal_dir: Union[pathlib.Path, str], split: str, pre_process: anomalib.pre_processing.PreProcessor, normal_test_dir: Optional[Union[pathlib.Path, str]] = None, split_ratio: float = 0.2, mask_dir: Optional[Union[pathlib.Path, str]] = None, extensions: Optional[Tuple[str, Ellipsis]] = None, task: Optional[str] = None, seed: int = 0, create_validation_set: bool = False)[source]

Bases: torch.utils.data.Dataset

Folder Dataset.

__len__(self) → int[source]

Get length of the dataset.

__getitem__(self, index: int) → Dict[str, Union[str, torch.Tensor]][source]

Get dataset item for the index index.

Parameters

index (int) – Index to get the item.

Returns

Dict of image tensor during training.

Otherwise, Dict containing image path, target path, image tensor, label and transformed bounding box.

Return type

Union[Dict[str, Tensor], Dict[str, Union[str, Tensor]]]

class anomalib.data.folder.Folder(root: Union[str, pathlib.Path], normal_dir: str = 'normal', abnormal_dir: str = 'abnormal', task: str = 'classification', normal_test_dir: Optional[Union[pathlib.Path, str]] = None, mask_dir: Optional[Union[pathlib.Path, str]] = None, extensions: Optional[Tuple[str, Ellipsis]] = None, split_ratio: float = 0.2, seed: int = 0, image_size: Optional[Union[int, Tuple[int, int]]] = None, train_batch_size: int = 32, test_batch_size: int = 32, num_workers: int = 8, transform_config_train: Optional[Union[str, albumentations.Compose]] = None, transform_config_val: Optional[Union[str, albumentations.Compose]] = None, create_validation_set: bool = False)[source]

Bases: pytorch_lightning.core.datamodule.LightningDataModule

Folder Lightning Data Module.

setup(self, stage: Optional[str] = None) → None[source]

Setup train, validation and test data.

Parameters

stage – Optional[str]: Train/Val/Test stages. (Default value = None)

train_dataloader(self) → pytorch_lightning.utilities.types.TRAIN_DATALOADERS[source]

Get train dataloader.

val_dataloader(self) → pytorch_lightning.utilities.types.EVAL_DATALOADERS[source]

Get validation dataloader.

test_dataloader(self) → pytorch_lightning.utilities.types.EVAL_DATALOADERS[source]

Get test dataloader.

predict_dataloader(self) → pytorch_lightning.utilities.types.EVAL_DATALOADERS[source]

Get predict dataloader.