RCAN: Image Super-Resolution Using Very DeepResidual Channel Attention Networks

RCAN是引用较高的超分模型。paper code
作者观察到两个问题：

1. 深层网络有训练难度。
2. LR中有低频成分，和相对高频且更重要的成分。但二者被平等对待。

针对这两点的解决思路：

1. 采用嵌套的残差结构加深网络，RIR residual in residua。
2. 在通道（不同成分）上增加注意力（权重）。

RCAN网络结构：
RCAB结构：

RCAB代码：

## Residual Channel Attention Block (RCAB)
class RCAB(nn.Module):
    def __init__(
        self, conv, n_feat, kernel_size, reduction,
        bias=True, bn=False, act=nn.ReLU(True), res_scale=1):

        super(RCAB, self).__init__()
        modules_body = []
        for i in range(2):
            modules_body.append(conv(n_feat, n_feat, kernel_size, bias=bias))
            if bn: modules_body.append(nn.BatchNorm2d(n_feat))
            if i == 0: modules_body.append(act)
        modules_body.append(CALayer(n_feat, reduction))
        self.body = nn.Sequential(*modules_body)
        self.res_scale = res_scale

    def forward(self, x):
        res = self.body(x)
        #res = self.body(x).mul(self.res_scale)
        res += x
        return res

channel attention：
CA代码：

class CALayer(nn.Module):
    def __init__(self, channel, reduction=16):
        super(CALayer, self).__init__()
        # global average pooling: feature --> point
        self.avg_pool = nn.AdaptiveAvgPool2d(1)
        # feature channel downscale and upscale --> channel weight
        self.conv_du = nn.Sequential(
                nn.Conv2d(channel, channel // reduction, 1, padding=0, bias=True),
                nn.ReLU(inplace=True),
                nn.Conv2d(channel // reduction, channel, 1, padding=0, bias=True),
                nn.Sigmoid()
        )

    def forward(self, x):
        y = self.avg_pool(x)
        y = self.conv_du(y)
        return x * y

思考：

1. 在训练初期，每个kernal还没有区分度，CA本身的学习和初始化参数是否会对网络参数学习产生影响？

参考文献：

1. https://blog.cython.top/paper-notes-1.html#architecture
2. https://www.kesci.com/home/project/5db292f875df5c002b23cda0
3. http://www.jeepxie.net/article/611013.html
4. https://perper.site/2019/09/05/Image-Super-Resolution-Using-Very-Deep-Residual-Channel-Attention-Networks-RCAN/
5. https://www.jianshu.com/p/07334544b4b0