spandrel.architectures.Compact.Compact

class documentation

class Compact(nn.Module): (source)

Constructor: Compact(num_in_ch, num_out_ch, num_feat, num_conv, ...)

View In Hierarchy

A compact VGG-style network structure for super-resolution.

It is a compact network structure, which performs upsampling in the last layer and no convolution is conducted on the HR feature space.

Parameters
num_in_chChannel number of inputs. Default: 3.
num_out_chChannel number of outputs. Default: 3.
num_featChannel number of intermediate features. Default: 64.
num_convNumber of convolution layers in the body network. Default: 16.
upscaleUpsampling factor. Default: 4.
act_typeActivation type, options: 'relu', 'prelu', 'leakyrelu'. Default: prelu.
Method __init__ Undocumented
Method forward Undocumented
Class Variable hyperparameters Undocumented
Instance Variable act_type Undocumented
Instance Variable body Undocumented
Instance Variable num_conv Undocumented
Instance Variable num_feat Undocumented
Instance Variable num_in_ch Undocumented
Instance Variable num_out_ch Undocumented
Instance Variable upsampler Undocumented
Instance Variable upscale Undocumented
def __init__(self, *, num_in_ch=3, num_out_ch=3, num_feat=64, num_conv=16, upscale=4, act_type='prelu'): (source)

Undocumented

def forward(self, x): (source)

Undocumented

hyperparameters: dict = (source)

Undocumented

act_type: str = (source)

Undocumented

body = (source)

Undocumented

num_conv: int = (source)

Undocumented

num_feat: int = (source)

Undocumented

num_in_ch: int = (source)

Undocumented

num_out_ch: int = (source)

Undocumented

upsampler = (source)

Undocumented

upscale: int = (source)

Undocumented