sup3r.utilities.loss_metrics.PerceptualLoss#

class PerceptualLoss(layer_names=None)[source]#

Perceptual loss that is calculated as MSE between feature maps of ground truth and synthetic data

Parameters:: layer_names (list | None) – List of layer names in VGG16 to use to extract feature maps from ground truth and synthetic data. Defaults to [‘block1_conv2’, ‘block2_conv2’]

Methods

`call`(x_true, x_gen)	Perceptual loss calculated on true and synthetic feature maps
`from_config`(config)
`get_config`()

Attributes

dtype

call(x_true, x_gen)[source]#

Perceptual loss calculated on true and synthetic feature maps

Parameters:

x_true (tf.tensor) – True high resolution data, shape is either of these: (n_obs, spatial_1, spatial_2, features) (n_obs, spatial_1, spatial_2, temporal, features)
x_gen (tf.tensor) – Synthetic high-res generator output, shape is either of these: (n_obs, spatial_1, spatial_2, features) (n_obs, spatial_1, spatial_2, temporal, features)

Returns:

tf.tensor – 0D tensor loss value

__call__(y_true, y_pred, sample_weight=None)#: Call self as a function.

sup3r.utilities.loss_metrics.PerceptualLoss