Total Variation (TV)

Module Interface

class torchmetrics.image.TotalVariation(reduction='sum', **kwargs)

Compute Total Variation loss (TV).

As input to forward and update the metric accepts the following input

• img (Tensor): A tensor of shape (N, C, H, W) consisting of images

As output of forward and compute the metric returns the following output

• sdi (Tensor): if reduction!='none' returns float scalar tensor with average TV value over sample else returns tensor of shape (N,) with TV values per sample

Parameters:

• reduction (Optional[Literal['mean', 'sum', 'none']]) –

  a method to reduce metric score over samples

  • 'mean': takes the mean over samples

  • 'sum': takes the sum over samples

  • None or 'none': return the score per sample

• kwargs (Any) – Additional keyword arguments, see Advanced metric settings for more info.

Raises:

ValueError – If reduction is not one of 'sum', 'mean', 'none' or None

Example

>>> import torch
>>> from torchmetrics.image import TotalVariation
>>> _ = torch.manual_seed(42)
>>> tv = TotalVariation()
>>> img = torch.rand(5, 3, 28, 28)
>>> tv(img)
tensor(7546.8018)

plot(val=None, ax=None)

Plot a single or multiple values from the metric.

Parameters:

• val (Union[Tensor, Sequence[Tensor], None]) – Either a single result from calling metric.forward or metric.compute or a list of these results. If no value is provided, will automatically call metric.compute and plot that result.

• ax (Optional[Axes]) – An matplotlib axis object. If provided will add plot to that axis

Return type:

Tuple[Figure, Union[Axes, ndarray]]

Returns:

Figure and Axes object

Raises:

ModuleNotFoundError – If matplotlib is not installed

>>> # Example plotting a single value
>>> import torch
>>> from torchmetrics.image import TotalVariation
>>> metric = TotalVariation()
>>> metric.update(torch.rand(5, 3, 28, 28))
>>> fig_, ax_ = metric.plot()

>>> # Example plotting multiple values
>>> import torch
>>> from torchmetrics.image import TotalVariation
>>> metric = TotalVariation()
>>> values = [ ]
>>> for _ in range(10):
...     values.append(metric(torch.rand(5, 3, 28, 28)))
>>> fig_, ax_ = metric.plot(values)

Functional Interface

torchmetrics.functional.image.total_variation(img, reduction='sum')

Compute total variation loss.

Parameters:

• img (Tensor) – A Tensor of shape (N, C, H, W) consisting of images

• reduction (Optional[Literal['mean', 'sum', 'none']]) –

  a method to reduce metric score over samples.

  • 'mean': takes the mean over samples

  • 'sum': takes the sum over samples

  • None or 'none': return the score per sample

Return type:

Tensor

Returns:

A loss scalar value containing the total variation

Raises:

• ValueError – If reduction is not one of 'sum', 'mean', 'none' or None

• RuntimeError – If img is not 4D tensor

Example

>>> import torch
>>> from torchmetrics.functional.image import total_variation
>>> _ = torch.manual_seed(42)
>>> img = torch.rand(5, 3, 28, 28)
>>> total_variation(img)
tensor(7546.8018)