ERGAS

class deepinv.loss.metric.ERGAS(factor: int, *args, **kwargs)

Bases: Metric

Error relative global dimensionless synthesis metric.

Calculates the ERGAS metric on a multispectral image and a target. ERGAS is a popular metric for pan-sharpening of multispectral images.

Wraps the torchmetrics ERGAS function. Note that our reduction parameter follows our uniform convention (see below).

Note

By default, no reduction is performed in the batch dimension.

Example:

>>> import torch
>>> from deepinv.loss.metric import ERGAS
>>> m = ERGAS(factor=4)
>>> x_net = x = torch.ones(3, 2, 8, 8) # B,C,H,W
>>> m(x_net, x)
tensor([0., 0., 0.])

Parameters:

• factor (int) – pansharpening factor.

• train_loss (bool) – use metric as a training loss, by returning one minus the metric.

• reduction (str) – a method to reduce metric score over individual batch scores. mean: takes the mean, sum takes the sum, none or None no reduction will be applied (default).

• norm_inputs (str) – normalize images before passing to metric. l2``normalizes by L2 spatial norm, ``min_max normalizes by min and max of each input.

Examples using ERGAS:

Remote sensing with satellite images

Remote sensing with satellite images