SameSizeTensorValues

class hybrid_learning.datasets.transforms.dict_transforms.SameSizeTensorValues(mode='up_bilinear')[source]

Bases: DictTransform

Up- or down-scale the tensor mask values of a dictionary to all have the same size. The mode determines whether and how it is up- or downscaled to the largest/smallest occurring size. Mask tensor sizes are interpreted as ([batch_size[, num_channels],] height, width). In case the mask has more than one channel, each channel is resized separately.

Public Data Attributes:

settings

Settings to reproduce the instance.
Inherited from : py: class:DictTransform

settings

Settings to reproduce the instance.
Inherited from : py: class:Transform

IDENTITY_CLASS

The identity class or classes for composition / addition.

settings

Settings to reproduce the instance.

Public Methods:

apply_to(annotations)

Up- or downscale the values of annotations to the same size according to mode.
Inherited from : py: class:DictTransform

apply_to(annotations)

Up- or downscale the values of annotations to the same size according to mode.
Inherited from : py: class:Transform

apply_to(annotations)

Up- or downscale the values of annotations to the same size according to mode.

Special Methods:

__init__([mode])

Inherited from : py: class:DictTransform

__call__(annotations)

Call method modifying a given dictionary.
Inherited from : py: class:Transform

__repr__()

Return repr(self).

__eq__(other)

Return self==value.

__copy__()

Return a shallow copy of self using settings.

__add__(other)

Return a flat composition of self with other.

__radd__(other)

Return a flat composition of other and self.

__call__(annotations)

Call method modifying a given dictionary.
Parameters

mode (Literal['up_bilinear', 'down_max']) –

__init__(mode='up_bilinear')[source]
Parameters

mode (Literal['up_bilinear', 'down_max']) –

apply_to(annotations)[source]

Up- or downscale the values of annotations to the same size according to mode. Note that for efficiency reasons annotations is modified in-place!

Parameters

annotations (MutableMapping[str, Any]) –

Return type

Union[Mapping[str, Any], Any]

mode: Literal['up_bilinear', 'down_max']

Up- or downscaling mode. Allowed values:

  • up_{interpolation}: upscaling to largest size using {interpolation};

    the interpolation mode must be one of the options for torch.nn.functional.interpolate(); equivalent to OnValues(Resize(max_size))

  • down_max: down-scaling via max-pooling

property settings: Dict[str, Any]

Settings to reproduce the instance.