2023-04-10 22:49:56 +00:00
|
|
|
# from https://github.com/cszn/BSRGAN/blob/main/models/network_rrdbnet.py
|
|
|
|
|
|
|
|
import functools
|
|
|
|
|
|
|
|
import torch
|
|
|
|
import torch.nn as nn
|
|
|
|
import torch.nn.functional as F
|
|
|
|
import torch.nn.init as init
|
|
|
|
|
|
|
|
|
|
|
|
def initialize_weights(net_l, scale=1):
|
|
|
|
if not isinstance(net_l, list):
|
|
|
|
net_l = [net_l]
|
|
|
|
for net in net_l:
|
|
|
|
for m in net.modules():
|
|
|
|
if isinstance(m, nn.Conv2d):
|
|
|
|
init.kaiming_normal_(m.weight, a=0, mode="fan_in")
|
|
|
|
m.weight.data *= scale # for residual block
|
|
|
|
if m.bias is not None:
|
|
|
|
m.bias.data.zero_()
|
|
|
|
elif isinstance(m, nn.Linear):
|
|
|
|
init.kaiming_normal_(m.weight, a=0, mode="fan_in")
|
|
|
|
m.weight.data *= scale
|
|
|
|
if m.bias is not None:
|
|
|
|
m.bias.data.zero_()
|
|
|
|
elif isinstance(m, nn.BatchNorm2d):
|
|
|
|
init.constant_(m.weight, 1)
|
|
|
|
init.constant_(m.bias.data, 0.0)
|
|
|
|
|
|
|
|
|
|
|
|
def make_layer(block, n_layers):
|
|
|
|
layers = []
|
|
|
|
for _ in range(n_layers):
|
|
|
|
layers.append(block())
|
|
|
|
return nn.Sequential(*layers)
|
|
|
|
|
|
|
|
|
|
|
|
class ResidualDenseBlock_5C(nn.Module):
|
|
|
|
def __init__(self, nf=64, gc=32, bias=True):
|
|
|
|
super(ResidualDenseBlock_5C, self).__init__()
|
|
|
|
# gc: growth channel, i.e. intermediate channels
|
|
|
|
self.conv1 = nn.Conv2d(nf, gc, 3, 1, 1, bias=bias)
|
|
|
|
self.conv2 = nn.Conv2d(nf + gc, gc, 3, 1, 1, bias=bias)
|
|
|
|
self.conv3 = nn.Conv2d(nf + 2 * gc, gc, 3, 1, 1, bias=bias)
|
|
|
|
self.conv4 = nn.Conv2d(nf + 3 * gc, gc, 3, 1, 1, bias=bias)
|
|
|
|
self.conv5 = nn.Conv2d(nf + 4 * gc, nf, 3, 1, 1, bias=bias)
|
|
|
|
self.lrelu = nn.LeakyReLU(negative_slope=0.2, inplace=True)
|
|
|
|
|
|
|
|
# initialization
|
|
|
|
initialize_weights(
|
|
|
|
[self.conv1, self.conv2, self.conv3, self.conv4, self.conv5], 0.1
|
|
|
|
)
|
|
|
|
|
|
|
|
def forward(self, x):
|
|
|
|
x1 = self.lrelu(self.conv1(x))
|
|
|
|
x2 = self.lrelu(self.conv2(torch.cat((x, x1), 1)))
|
|
|
|
x3 = self.lrelu(self.conv3(torch.cat((x, x1, x2), 1)))
|
|
|
|
x4 = self.lrelu(self.conv4(torch.cat((x, x1, x2, x3), 1)))
|
|
|
|
x5 = self.conv5(torch.cat((x, x1, x2, x3, x4), 1))
|
|
|
|
return x5 * 0.2 + x
|
|
|
|
|
|
|
|
|
|
|
|
class RRDB(nn.Module):
|
|
|
|
"""Residual in Residual Dense Block"""
|
|
|
|
|
|
|
|
def __init__(self, nf, gc=32):
|
|
|
|
super(RRDB, self).__init__()
|
|
|
|
self.RDB1 = ResidualDenseBlock_5C(nf, gc)
|
|
|
|
self.RDB2 = ResidualDenseBlock_5C(nf, gc)
|
|
|
|
self.RDB3 = ResidualDenseBlock_5C(nf, gc)
|
|
|
|
|
|
|
|
def forward(self, x):
|
|
|
|
out = self.RDB1(x)
|
|
|
|
out = self.RDB2(out)
|
|
|
|
out = self.RDB3(out)
|
|
|
|
return out * 0.2 + x
|
|
|
|
|
|
|
|
|
|
|
|
class RRDBNet(nn.Module):
|
2023-04-15 21:55:53 +00:00
|
|
|
def __init__(
|
|
|
|
self,
|
|
|
|
num_in_ch=3,
|
|
|
|
num_out_ch=3,
|
|
|
|
num_feat=64,
|
|
|
|
num_block=23,
|
|
|
|
num_grow_ch=32,
|
|
|
|
scale=4,
|
|
|
|
):
|
2023-04-10 22:49:56 +00:00
|
|
|
super(RRDBNet, self).__init__()
|
2023-04-15 21:20:46 +00:00
|
|
|
RRDB_block_f = functools.partial(RRDB, nf=num_feat, gc=num_grow_ch)
|
|
|
|
self.sf = scale
|
|
|
|
print([num_in_ch, num_out_ch, num_feat, num_block, num_grow_ch, scale])
|
2023-04-10 22:49:56 +00:00
|
|
|
|
2023-04-15 21:20:46 +00:00
|
|
|
self.conv_first = nn.Conv2d(num_in_ch, num_feat, 3, 1, 1, bias=True)
|
|
|
|
self.RRDB_trunk = make_layer(RRDB_block_f, num_block)
|
|
|
|
self.trunk_conv = nn.Conv2d(num_feat, num_feat, 3, 1, 1, bias=True)
|
2023-04-10 22:49:56 +00:00
|
|
|
# upsampling
|
2023-04-15 21:20:46 +00:00
|
|
|
self.upconv1 = nn.Conv2d(num_feat, num_feat, 3, 1, 1, bias=True)
|
2023-04-10 22:49:56 +00:00
|
|
|
if self.sf == 4:
|
2023-04-15 21:20:46 +00:00
|
|
|
self.upconv2 = nn.Conv2d(num_feat, num_feat, 3, 1, 1, bias=True)
|
|
|
|
self.HRconv = nn.Conv2d(num_feat, num_feat, 3, 1, 1, bias=True)
|
|
|
|
self.conv_last = nn.Conv2d(num_feat, num_out_ch, 3, 1, 1, bias=True)
|
2023-04-10 22:49:56 +00:00
|
|
|
|
|
|
|
self.lrelu = nn.LeakyReLU(negative_slope=0.2, inplace=True)
|
|
|
|
|
|
|
|
def forward(self, x):
|
|
|
|
fea = self.conv_first(x)
|
|
|
|
trunk = self.trunk_conv(self.RRDB_trunk(fea))
|
|
|
|
fea = fea + trunk
|
|
|
|
|
|
|
|
fea = self.lrelu(
|
|
|
|
self.upconv1(F.interpolate(fea, scale_factor=2, mode="nearest"))
|
|
|
|
)
|
|
|
|
if self.sf == 4:
|
|
|
|
fea = self.lrelu(
|
|
|
|
self.upconv2(F.interpolate(fea, scale_factor=2, mode="nearest"))
|
|
|
|
)
|
|
|
|
out = self.conv_last(self.lrelu(self.HRconv(fea)))
|
|
|
|
|
|
|
|
return out
|