Frechet Inception Distance (FID)

Module Interface

class torchmetrics.image.fid.FrechetInceptionDistance(feature=2048, reset_real_features=True, normalize=False, input_img_size=(3, 299, 299), **kwargs)

Calculate Fréchet inception distance (FID) which is used to access the quality of generated images.

\[FID = \|\mu - \mu_w\|^2 + tr(\Sigma + \Sigma_w - 2(\Sigma \Sigma_w)^{\frac{1}{2}})\]

where \(\mathcal{N}(\mu, \Sigma)\) is the multivariate normal distribution estimated from Inception v3 (fid ref1) features calculated on real life images and \(\mathcal{N}(\mu_w, \Sigma_w)\) is the multivariate normal distribution estimated from Inception v3 features calculated on generated (fake) images. The metric was originally proposed in fid ref1.

Using the default feature extraction (Inception v3 using the original weights from fid ref2), the input is expected to be mini-batches of 3-channel RGB images of shape (3xHxW). If argument normalize is True images are expected to be dtype float and have values in the [0,1] range, else if normalize is set to False images are expected to have dtype uint8 and take values in the [0, 255] range. All images will be resized to 299 x 299 which is the size of the original training data. The boolian flag real determines if the images should update the statistics of the real distribution or the fake distribution.

Using custom feature extractor is also possible. One can give a torch.nn.Module as feature argument. This custom feature extractor is expected to have output shape of (1, num_features). This would change the used feature extractor from default (Inception v3) to the given network. In case network doesn’t have num_features attribute, a random tensor will be given to the network to infer feature dimensionality. Size of this tensor can be controlled by input_img_size argument and type of the tensor can be controlled with normalize argument (True uses float32 tensors and False uses int8 tensors). In this case, update method expects to have the tensor given to imgs argument to be in the correct shape and type that is compatible to the custom feature extractor.

This metric is known to be unstable in its calculatations, and we recommend for the best results using this metric that you calculate using torch.float64 (default is torch.float32) which can be set using the .set_dtype method of the metric.

Hint

Using this metric with the default feature extractor requires that torch-fidelity is installed. Either install as pip install torchmetrics[image] or pip install torch-fidelity

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

• imgs (Tensor): tensor with images feed to the feature extractor with

• real (bool): bool indicating if imgs belong to the real or the fake distribution

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

• fid (Tensor): float scalar tensor with mean FID value over samples

Parameters:

• feature (Union[int, Module]) –

  Either an integer or nn.Module:

  • an integer will indicate the inceptionv3 feature layer to choose. Can be one of the following: 64, 192, 768, 2048

  • an nn.Module for using a custom feature extractor. Expects that its forward method returns an (N,d) matrix where N is the batch size and d is the feature size.

• reset_real_features (bool) – Whether to also reset the real features. Since in many cases the real dataset does not change, the features can be cached them to avoid recomputing them which is costly. Set this to False if your dataset does not change.

• normalize (bool) –

  Argument for controlling the input image dtype normalization:

  • If default feature extractor is used, controls whether input imgs have values in range [0, 1] or not:

    • True: if input imgs have values ranged in [0, 1]. They are cast to int8/byte tensors.

    • False: if input imgs have values ranged in [0, 255]. No casting is done.

  • If custom feature extractor module is used, controls type of the input img tensors:

    • True: if input imgs are expected to be in the data type of torch.float32.

    • False: if input imgs are expected to be in the data type of torch.int8.

• input_img_size (Tuple[int, int, int]) – tuple of integers. Indicates input img size to the custom feature extractor network if provided.

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

Raises:

• ValueError – If torch version is lower than 1.9

• ModuleNotFoundError – If feature is set to an int (default settings) and torch-fidelity is not installed

• ValueError – If feature is set to an int not in [64, 192, 768, 2048]

• TypeError – If feature is not an str, int or torch.nn.Module

• ValueError – If reset_real_features is not an bool

Example

>>> from torch import rand
>>> from torchmetrics.image.fid import FrechetInceptionDistance
>>> fid = FrechetInceptionDistance(feature=64)
>>> # generate two slightly overlapping image intensity distributions
>>> imgs_dist1 = torch.randint(0, 200, (100, 3, 299, 299), dtype=torch.uint8)
>>> imgs_dist2 = torch.randint(100, 255, (100, 3, 299, 299), dtype=torch.uint8)
>>> fid.update(imgs_dist1, real=True)
>>> fid.update(imgs_dist2, real=False)
>>> fid.compute()
tensor(12.6388)

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.fid import FrechetInceptionDistance
>>> imgs_dist1 = torch.randint(0, 200, (100, 3, 299, 299), dtype=torch.uint8)
>>> imgs_dist2 = torch.randint(100, 255, (100, 3, 299, 299), dtype=torch.uint8)
>>> metric = FrechetInceptionDistance(feature=64)
>>> metric.update(imgs_dist1, real=True)
>>> metric.update(imgs_dist2, real=False)
>>> fig_, ax_ = metric.plot()

>>> # Example plotting multiple values
>>> import torch
>>> from torchmetrics.image.fid import FrechetInceptionDistance
>>> imgs_dist1 = lambda: torch.randint(0, 200, (100, 3, 299, 299), dtype=torch.uint8)
>>> imgs_dist2 = lambda: torch.randint(100, 255, (100, 3, 299, 299), dtype=torch.uint8)
>>> metric = FrechetInceptionDistance(feature=64)
>>> values = [ ]
>>> for _ in range(3):
...     metric.update(imgs_dist1(), real=True)
...     metric.update(imgs_dist2(), real=False)
...     values.append(metric.compute())
...     metric.reset()
>>> fig_, ax_ = metric.plot(values)

reset()

Reset metric states.

Return type:

None

set_dtype(dst_type)

Transfer all metric state to specific dtype. Special version of standard type method.

Parameters:

dst_type (Union[str, dtype]) – the desired type as torch.dtype or string

Return type:

Metric