FID#
- class ignite.metrics.FID(num_features=None, feature_extractor=None, output_transform=<function FID.<lambda>>, device=device(type='cpu'))[source]#
Calculates Frechet Inception Distance.
$\text{FID} = |\mu_{1} - \mu_{2}| + \text{Tr}(\sigma_{1} + \sigma_{2} - {2}\sqrt{\sigma_1*\sigma_2})$where $\mu_1$ and $\sigma_1$ refer to the mean and covariance of the train data and $\mu_2$ and $\sigma_2$ refer to the mean and covariance of the test data.
More details can be found in Heusel et al. 2002
In addition, a faster and online computation approach can be found in Chen et al. 2014
Remark:
This implementation is inspired by pytorch_fid package which can be found here
Note
The default Inception model requires the torchvision module to be installed. FID also requires scipy library for matrix square root calculations.
- Parameters
num_features (Optional[int]) – number of features predicted by the model or the reduced feature vector of the image. Default value is 2048.
feature_extractor (Optional[torch.nn.modules.module.Module]) – a torch Module for extracting the features from the input data. It returns a tensor of shape (batch_size, num_features). If neither
num_features
norfeature_extractor
are defined, by default we use an ImageNet pretrained Inception Model. If onlynum_features
is defined butfeature_extractor
is not defined,feature_extractor
is assigned Identity Function. Please note that the model will be implicitly converted to device mentioned in thedevice
argument.output_transform (Callable) – a callable that is used to transform the
Engine
’sprocess_function
’s output into the form expected by the metric. This can be useful if, for example, you have a multi-output model and you want to compute the metric with respect to one of the outputs. By default, metrics require the output as(y_pred, y)
or{'y_pred': y_pred, 'y': y}
.device (Union[str, torch.device]) – specifies which device updates are accumulated on. Setting the metric’s device to be the same as your
update
arguments ensures theupdate
method is non-blocking. By default, CPU.
- Return type
Examples
metric = FID() metric.attach(default_evaluator, "fid") y_true = torch.rand(10, 3, 299, 299) y_pred = torch.rand(10, 3, 299, 299) state = default_evaluator.run([[y_pred, y_true]]) print(state.metrics["fid"])
metric = FID(num_features=1, feature_extractor=default_model) metric.attach(default_evaluator, "fid") y_true = torch.ones(10, 4) y_pred = torch.ones(10, 4) state = default_evaluator.run([[y_pred, y_true]]) print(state.metrics["fid"])
0.0
New in version 0.4.6.
Methods
Computes the metric based on it's accumulated state.
Resets the metric to it's initial state.
Updates the metric's state using the passed batch output.
- compute()[source]#
Computes the metric based on it’s accumulated state.
By default, this is called at the end of each epoch.
- Returns
- the actual quantity of interest. However, if a
Mapping
is returned, it will be (shallow) flattened into engine.state.metrics whencompleted()
is called. - Return type
Any
- Raises
NotComputableError – raised when the metric cannot be computed.
- reset()[source]#
Resets the metric to it’s initial state.
By default, this is called at the start of each epoch.
- Return type
- update(output)[source]#
Updates the metric’s state using the passed batch output.
By default, this is called once for each batch.
- Parameters
output (Sequence[torch.Tensor]) – the is the output from the engine’s process function.
- Return type