kaira.losses.image.L1Loss

Inheritance diagram of L1Loss — Inheritance diagram for L1Loss

class kaira.losses.image.L1Loss[source]

Bases: BaseLoss

L1 (Mean Absolute Error) Loss Module.

This module calculates the L1 loss between the input and the target. L1 loss is often preferred over MSE for image restoration tasks as it preserves edges better and is more robust to outliers [Zhao et al., 2016].

Methods

`__init__`	Initialize the L1Loss module.
`forward`	Forward pass through the L1Loss module.

__init__()[source]: Initialize the L1Loss module.

forward(x: Tensor, target: Tensor) → Tensor[source]

Forward pass through the L1Loss module.

Parameters:

x (torch.Tensor) – The input tensor.
target (torch.Tensor) – The target tensor.

Returns:

The L1 loss between the input and the target.

Return type:

torch.Tensor