定义卷积层
def default_conv(in_channels, out_channels, kernel_size, bias=True):return nn.Conv2d(in_channels, out_channels, kernel_size,padding=(kernel_size//2), bias=bias)对应：We pre-process all the images by subtracting the mean RGB value of the DIV2K dataset.
class MeanShift(nn.Conv2d):def __init__(self, rgb_range,rgb_mean=(0.4488, 0.4371, 0.4040), rgb_std=(1.0, 1.0, 1.0), sign=-1):super(MeanShift, self).__init__(3, 3, kernel_size=1)std = torch.Tensor(rgb_std)self.weight.data = torch.eye(3).view(3, 3, 1, 1) / std.view(3, 1, 1, 1)self.bias.data = sign * rgb_range * torch.Tensor(rgb_mean) / stdfor p in self.parameters():p.requires_grad = False定义 conv + bn + relu 模块
class BasicBlock(nn.Sequential):def __init__(self, conv, in_channels, out_channels, kernel_size, stride=1, bias=False,bn=True, act=nn.ReLU(True)):m = [conv(in_channels, out_channels, kernel_size, bias=bias)]if bn:m.append(nn.BatchNorm2d(out_channels))if act is not None:m.append(act)super(BasicBlock, self).__init__(*m)定义EDSR残差块
class ResBlock(nn.Module):def __init__(self, conv, n_feats, kernel_size,bias=True, bn=False, act=nn.ReLU(True), res_scale=1):super(ResBlock, self).__init__()m = []for i in range(2): # i = 0,1 m.append(conv(n_feats, n_feats, kernel_size, bias=bias))if bn:m.append(nn.BatchNorm2d(n_feats))if i == 0:m.append(act)self.body = nn.Sequential(*m)self.res_scale = res_scaledef forward(self, x):res = self.body(x).mul(self.res_scale)
#res乘以res_scale（the residual scaling with factor 0.1）达到constant scaling layers
#对应论文We resolve this issue by adopting the residual scaling with factor 0.1. In each #residual block, constant scaling layers are placed after the last convolution layers.res += xreturn res对应论文 conv+shuffle
class Upsampler(nn.Sequential):def __init__(self, conv, scale, n_feats, bn=False, act=False, bias=True):m = []if (scale & (scale - 1)) == 0: # Is scale = 2^n?for _ in range(int(math.log(scale, 2))):m.append(conv(n_feats, 4 * n_feats, 3, bias))m.append(nn.PixelShuffle(2))if bn:m.append(nn.BatchNorm2d(n_feats))if act == 'relu':m.append(nn.ReLU(True))elif act == 'prelu':m.append(nn.PReLU(n_feats))elif scale == 3:m.append(conv(n_feats, 9 * n_feats, 3, bias))m.append(nn.PixelShuffle(3))if bn:m.append(nn.BatchNorm2d(n_feats))if act == 'relu':m.append(nn.ReLU(True))elif act == 'prelu':m.append(nn.PReLU(n_feats))else:raise NotImplementedErrorsuper(Upsampler, self).__init__(*m)