Ctrl+K

Synthetic Data Utils

Synthetic Data Utils#

Utilities to generate synthetic data.

anomalib.data.utils.generators.random_2d_perlin(shape, res, fade=<function <lambda>>)#

Returns a random 2d perlin noise array.

Parameters:

  • shape (tuple) – Shape of the 2d map.

  • res (tuple[int | torch.Tensor, int | torch.Tensor]) – Tuple of scales for perlin noise for height and width dimension.

  • fade (_type_, optional) – Function used for fading the resulting 2d map. Defaults to equation 6*t**5-15*t**4+10*t**3.

Returns:

Random 2d-array/tensor generated using perlin noise.

Return type:

np.ndarray | torch.Tensor

Augmenter module to generates out-of-distribution samples for the DRAEM implementation.

class anomalib.data.utils.augmenter.Augmenter(anomaly_source_path=None, p_anomalous=0.5, beta=(0.2, 1.0))#

Bases: object

Class that generates noisy augmentations of input images.

Parameters:

  • anomaly_source_path (str | None) – Path to a folder of images that will be used as source of the anomalous

  • specified (noise. If not)

  • instead. (random noise will be used)

  • p_anomalous (float) – Probability that the anomalous perturbation will be applied to a given image.

  • beta (float) – Parameter that determines the opacity of the noise mask.

augment_batch(batch)#

Generate anomalous augmentations for a batch of input images.

Parameters:

batch (torch.Tensor) – Batch of input images

Return type:

tuple[Tensor, Tensor]

Returns:

  • Augmented image to which anomalous perturbations have been added.

  • Ground truth masks corresponding to the anomalous perturbations.

generate_perturbation(height, width, anomaly_source_path=None)#

Generate an image containing a random anomalous perturbation using a source image.

Parameters:

  • height (int) – height of the generated image.

  • width (int) – (int): width of the generated image.

  • anomaly_source_path (Path | str | None) – Path to an image file. If not provided, random noise will be used

  • instead.

Return type:

tuple[ndarray, ndarray]

Returns:

Image containing a random anomalous perturbation, and the corresponding ground truth anomaly mask.

rand_augmenter()#

Select 3 random transforms that will be applied to the anomaly source images.

Return type:

Sequential

Returns:

A selection of 3 transforms.

anomalib.data.utils.augmenter.nextpow2(value)#

Return the smallest power of 2 greater than or equal to the input value.

Return type:

int

Dataset that generates synthetic anomalies.

This dataset can be used when there is a lack of real anomalous data.

class anomalib.data.utils.synthetic.SyntheticAnomalyDataset(task, transform, source_samples)#

Bases: AnomalibDataset

Dataset which reads synthetically generated anomalous images from a temporary folder.

Parameters:

  • task (str) – Task type, either “classification” or “segmentation”.

  • transform (A.Compose) – Transform object describing the transforms that are applied to the inputs.

  • source_samples (DataFrame) – Normal samples to which the anomalous augmentations will be applied.

classmethod from_dataset(dataset)#

Create a synthetic anomaly dataset from an existing dataset of normal images.

Parameters:

dataset (AnomalibDataset) – Dataset consisting of only normal images that will be converrted to a synthetic anomalous dataset with a 50/50 normal anomalous split.

Return type:

SyntheticAnomalyDataset

anomalib.data.utils.synthetic.make_synthetic_dataset(source_samples, image_dir, mask_dir, anomalous_ratio=0.5)#

Convert a set of normal samples into a mixed set of normal and synthetic anomalous samples.

The synthetic images will be saved to the file system in the specified root directory under <root>/images. For the synthetic anomalous images, the masks will be saved under <root>/ground_truth.

Parameters:

  • source_samples (DataFrame) – Normal images that will be used as source for the synthetic anomalous images.

  • image_dir (Path) – Directory to which the synthetic anomalous image files will be written.

  • mask_dir (Path) – Directory to which the ground truth anomaly masks will be written.

  • anomalous_ratio (float) – Fraction of source samples that will be converted into anomalous samples.

Return type:

DataFrame

Contents