trw.transforms

This module is dedicated to data augmentations. In particular we strive to have a numpy and pytorch implementation for each augmentation so that we could if perform it on GPU

Transforms are designed to work for n-dimensional data.

Submodules

Package Contents

Classes

Transform

Abstraction of a batch transform

TransformBatchWithCriteria

Helper function to apply a given transform function on features that satisfy a criteria

TransformRandomCrop

Add padding on a numpy array of samples and random crop to original size

TransformRandomCropJoint

Add random padding & cropping on a numpy or Torch arrays. The arrays are joints and the same padding/cropping applied on all the arrays

TransformRandomFlip

Randomly flip the axis of selected features

TransformRandomFlipJoint

Randomly flip the axis of selected features in a joint fashion (if a feature is selected and a sample is

TransformRandomCutout

Randomly flip the axis of selected features

TransformResize

Resize a tensor to a fixed size

TransformNormalize

Normalize a tensor image with mean and standard deviation.

TransformCompose

Sequentially apply a list of transformations

Functions

transform_batch_random_crop(array, crop_shape, offsets=None, return_offsets=False)

Randomly crop a numpy array of samples given a target size. This works for an arbitrary number of dimensions

transform_batch_pad_numpy(array, padding, mode='edge', constant_value=0)

Add padding on a numpy array of samples. This works for an arbitrary number of dimensions

transform_batch_pad_torch(array, padding, mode='edge', constant_value=0)

Add padding on a numpy array of samples. This works for an arbitrary number of dimensions

transform_batch_pad(array, padding, mode='edge', constant_value=0)

Add padding on a numpy array of samples. This works for an arbitrary number of dimensions

flip

copy

cutout(image, cutout_size, cutout_value_fn)

Remove a part of the image randomly

resize

stack

normalize

renormalize(data, desired_mean, desired_std, current_mean=None, current_std=None)

Transform the data so that it has desired mean and standard deviation element wise

criteria_feature_name(feature_name, feature_value, feature_names)

Return True if the feature name belongs to a given set of names

criteria_is_array_3_or_above(feature_name, feature_value)

Return True if the feature is a numpy or torch array dim >= 3
trw.transforms.transform_batch_random_crop(array, crop_shape, offsets=None, return_offsets=False)

Randomly crop a numpy array of samples given a target size. This works for an arbitrary number of dimensions

Parameters

  • array – a numpy or Torch array. Samples are stored in the first dimension

  • crop_shape – a sequence of size len(array.shape)-1 indicating the shape of the crop

  • offsets – if None, offsets will be randomly created to crop with crop_shape, else an array indicating the crop position for each sample

  • return_offsets – if True, returns a tuple (cropped array, offsets)

Returns

a cropped array

trw.transforms.transform_batch_pad_numpy(array, padding, mode='edge', constant_value=0)

Add padding on a numpy array of samples. This works for an arbitrary number of dimensions

Parameters

  • array – a numpy array. Samples are stored in the first dimension

  • padding – a sequence of size len(array.shape)-1 indicating the width of the padding to be added at the beginning and at the end of each dimension (except for dimension 0)

  • mode – numpy.pad mode

Returns

a padded array

trw.transforms.transform_batch_pad_torch(array, padding, mode='edge', constant_value=0)

Add padding on a numpy array of samples. This works for an arbitrary number of dimensions

This function mimics the API of transform_batch_pad_numpy so they can be easily interchanged.

Parameters

  • array – a Torch array. Samples are stored in the first dimension

  • padding – a sequence of size len(array.shape)-1 indicating the width of the padding to be added at the beginning and at the end of each dimension (except for dimension 0)

  • mode – numpy.pad mode. Currently supported are (‘constant’, ‘edge’, ‘symmetric’)

Returns

a padded array

trw.transforms.transform_batch_pad(array, padding, mode='edge', constant_value=0)

Add padding on a numpy array of samples. This works for an arbitrary number of dimensions

Parameters

  • array – a numpy array. Samples are stored in the first dimension

  • padding – a sequence of size len(array.shape)-1 indicating the width of the padding to be added at the beginning and at the end of each dimension (except for dimension 0)

  • mode – numpy.pad mode

Returns

a padded array

trw.transforms.flip(array, axis)

Flip an axis of an array

Parameters

  • array – a numpy.ndarray or torch.Tensor n-dimensional array

  • axis – the xis to flip

Returns

an array with specified axis flipped

trw.transforms.copy(array)

Copy an array

Parameters

array – a numpy.ndarray or torch.Tensor n-dimensional array

Returns

an array with specified axis flipped

trw.transforms.cutout(image, cutout_size, cutout_value_fn)

Remove a part of the image randomly

Parameters

  • array – a numpy.ndarray or torch.Tensor n-dimensional array. Samples are stored on axis 0

  • cutout_size – the cutout_size of the regions to be occluded

  • cutout_value_fn – the function value used for occlusion. Must take as argument image and modify directly the image

Returns

None

trw.transforms.resize(array, size, mode='linear')

Resize the array

Parameters

  • array – a N-dimensional tensor, representing 1D to 3D data (3 to 5 dimensional data with dim 0 for the samples and dim 1 for filters)

  • size – a (N-2) list to which the array will be upsampled or downsampled

  • mode – string among (‘nearest’, ‘linear’) specifying the resampling method

Returns

a resized N-dimensional tensor

trw.transforms.stack(sequence, axis=0)

stack an array

Parameters

  • sequence – a numpy.ndarray or torch.Tensor n-dimensional array

  • axis – the xis to flip

Returns

an array stacked

trw.transforms.normalize(array, mean, std)

Normalize a tensor image with mean and standard deviation.

Given mean: (M1,…,Mn) and std: (S1,..,Sn) for n channels, this transform will normalize each channel of the input torch.Tensor, input[channel] = (input[channel] - mean[channel]) / std[channel]

Parameters

  • array – the torch array to normalize. Expected layout is (sample, filter, d0, … dN)

  • mean – a N-dimensional sequence

  • std – a N-dimensional sequence

Returns

A normalized tensor such that the mean is 0 and std is 1

trw.transforms.renormalize(data, desired_mean, desired_std, current_mean=None, current_std=None)

Transform the data so that it has desired mean and standard deviation element wise

Parameters

  • data – a torch or numpy array

  • desired_mean – the mean to transform data to

  • desired_std – the std to transform data to

  • current_mean – if the mean if known, do not recalculate it (e.g., training mean to be used in validation split)

  • current_std – if the std if known, do not recalculate it (e.g., training std to be used in validation split)

Returns

a data with mean desired_mean and std desired_std

class trw.transforms.Transform

Abstraction of a batch transform

abstract __call__(self, batch)
class trw.transforms.TransformBatchWithCriteria(criteria_fn, transform_fn)

Bases: Transform

Helper function to apply a given transform function on features that satisfy a criteria

__call__(self, batch)
trw.transforms.criteria_feature_name(feature_name, feature_value, feature_names)

Return True if the feature name belongs to a given set of names

trw.transforms.criteria_is_array_3_or_above(feature_name, feature_value)

Return True if the feature is a numpy or torch array dim >= 3

class trw.transforms.TransformRandomCrop(padding, criteria_fn=None, mode='edge', constant_value=0, size=None)

Bases: trw.transforms.transforms.TransformBatchWithCriteria

Add padding on a numpy array of samples and random crop to original size

Parameters

  • padding – a sequence of size len(array.shape)-1 indicating the width of the padding to be added at the beginning and at the end of each dimension (except for dimension 0). If None, no padding added

  • criteria_fn – function applied on each feature. If satisfied, the feature will be transformed, if not the original feature is returned

  • mode – numpy.pad mode. Currently supported are (‘constant’, ‘edge’, ‘symmetric’)

  • size – the size of the cropped image. If None, same size as input image

Returns

a randomly cropped batch

class trw.transforms.TransformRandomCropJoint(feature_names, padding, mode='edge', constant_value=0, size=None)

Bases: trw.transforms.transforms.TransformBatchJointWithCriteria

Add random padding & cropping on a numpy or Torch arrays. The arrays are joints and the same padding/cropping applied on all the arrays

Parameters

  • feature_names – these are the features that will be jointly padded and cropped

  • padding – a sequence of size len(array.shape)-1 indicating the width of the padding to be added at the beginning and at the end of each dimension (except for dimension 0). If None, no padding added

  • criteria_fn – function applied on each feature. If satisfied, the feature will be transformed, if not the original feature is returned

  • mode – numpy.pad mode. Currently supported are (‘constant’, ‘edge’, ‘symmetric’)

  • size – the size of the cropped image. If None, same size as input image

Returns

a randomly cropped batch

class trw.transforms.TransformRandomFlip(axis, flip_probability=0.5, criteria_fn=None)

Bases: trw.transforms.transforms.TransformBatchWithCriteria

Randomly flip the axis of selected features

class trw.transforms.TransformRandomFlipJoint(feature_names, axis, flip_probability=0.5)

Bases: trw.transforms.transforms.TransformBatchWithCriteria

Randomly flip the axis of selected features in a joint fashion (if a feature is selected and a sample is

flipped, it will be flipped for all selected features)

class trw.transforms.TransformRandomCutout(cutout_size, criteria_fn=None, cutout_value_fn=functools.partial(cutout_function.cutout_value_fn_constant, value=0))

Bases: trw.transforms.transforms.TransformBatchWithCriteria

Randomly flip the axis of selected features

class trw.transforms.TransformResize(size, criteria_fn=None, mode='linear')

Bases: trw.transforms.transforms.TransformBatchWithCriteria

Resize a tensor to a fixed size

class trw.transforms.TransformNormalize(mean, std, criteria_fn=None)

Bases: trw.transforms.transforms.TransformBatchWithCriteria

Normalize a tensor image with mean and standard deviation.

Given mean: (M1,…,Mn) and std: (S1,..,Sn) for n channels, this transform will normalize each channel of the input torch.Tensor, input[channel] = (input[channel] - mean[channel]) / std[channel]

Parameters

  • array – the torch array to normalize. Expected layout is (sample, filter, d0, … dN)

  • mean – a N-dimensional sequence

  • std – a N-dimensional sequence

  • criteria_fn – function applied on each feature. If satisfied, the feature will be transformed, if not the original feature is returned

Returns

A normalized batch such that the mean is 0 and std is 1 for the selected features

class trw.transforms.TransformCompose(transforms)

Bases: trw.transforms.transforms.Transform

Sequentially apply a list of transformations

__call__(self, batch)