basicsr.archs.ridnet_arch

class basicsr.archs.ridnet_arch.ChannelAttention(mid_channels, squeeze_factor=16)

Bases: Module

Channel attention.

Parameters:

• num_feat (int) – Channel number of intermediate features.

• squeeze_factor (int) – Channel squeeze factor. Default:

forward(x)

Define the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.

training: bool

class basicsr.archs.ridnet_arch.EAM(in_channels, mid_channels, out_channels)

Bases: Module

Enhancement attention modules (EAM) in RIDNet.

This module contains a merge-and-run unit, a residual block, an enhanced residual block and a feature attention unit.

merge

The merge-and-run unit.

block1

The residual block.

block2

The enhanced residual block.

ca

The feature/channel attention unit.

forward(x)

Define the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.

training: bool

class basicsr.archs.ridnet_arch.EResidualBlockNoBN(in_channels, out_channels)

Bases: Module

Enhanced Residual block without BN.

There are three convolution layers in residual branch.

forward(x)

Define the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.

training: bool

class basicsr.archs.ridnet_arch.MeanShift(rgb_range, rgb_mean, rgb_std, sign=-1, requires_grad=True)

Bases: Conv2d

Data normalization with mean and std.

Parameters:

• rgb_range (int) – Maximum value of RGB.

• rgb_mean (list[float]) – Mean for RGB channels.

• rgb_std (list[float]) – Std for RGB channels.

• sign (int) – For subtraction, sign is -1, for addition, sign is 1. Default: -1.

• requires_grad (bool) – Whether to update the self.weight and self.bias. Default: True.

bias: Tensor | None

dilation: Tuple[int, ...]

groups: int

in_channels: int

kernel_size: Tuple[int, ...]

out_channels: int

output_padding: Tuple[int, ...]

padding: str | Tuple[int, ...]

padding_mode: str

stride: Tuple[int, ...]

transposed: bool

weight: Tensor

class basicsr.archs.ridnet_arch.MergeRun(in_channels, out_channels, kernel_size=3, stride=1, padding=1)

Bases: Module

Merge-and-run unit.

This unit contains two branches with different dilated convolutions, followed by a convolution to process the concatenated features.

Paper: Real Image Denoising with Feature Attention Ref git repo: https://github.com/saeed-anwar/RIDNet

forward(x)

Define the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.

training: bool

class basicsr.archs.ridnet_arch.RIDNet(in_channels, mid_channels, out_channels, num_block=4, img_range=255.0, rgb_mean=(0.4488, 0.4371, 0.404), rgb_std=(1.0, 1.0, 1.0))

Bases: Module

RIDNet: Real Image Denoising with Feature Attention.

Ref git repo: https://github.com/saeed-anwar/RIDNet

Parameters:

• in_channels (int) – Channel number of inputs.

• mid_channels (int) – Channel number of EAM modules. Default: 64.

• out_channels (int) – Channel number of outputs.

• num_block (int) – Number of EAM. Default: 4.

• img_range (float) – Image range. Default: 255.

• rgb_mean (tuple[float]) – Image mean in RGB orders. Default: (0.4488, 0.4371, 0.4040), calculated from DIV2K dataset.

forward(x)

Define the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.

training: bool