语义分割中的nonlocal[3]-CCNet

ccnet是对non-local的加速，non-local中每个位置都会计算与全局的关系，ccnet 通过多次计算当前位置与其同行或者同列的feature的关系，逐步propagation到全局。

其中criss-cross attention block的结构如下：

多次叠加使用时信息的传递方式：

蓝色位置的信息第一次loop的时候被传递给浅绿色的位置，第二次loop的时候这些信息传递给了深绿色位置。

由于每个点只关注与（H+W-1）个点的关系，所以计算量会少。

另外叠加criss-cross attention次数越多，效果越好

ccnet的实现如下

def INF(B,H,W):
     return -torch.diag(torch.tensor(float("inf")).cuda().repeat(H),0).unsqueeze(0).repeat(B*W,1,1)


class CrissCrossAttention(nn.Module):
    # Criss-Cross Attention Module
    def __init__(self, in_dim):
        super(CrissCrossAttention,self).__init__()
        self.query_conv = nn.Conv2d(in_channels=in_dim, out_channels=in_dim//8, kernel_size=1)
        self.key_conv = nn.Conv2d(in_channels=in_dim, out_channels=in_dim//8, kernel_size=1)
        self.value_conv = nn.Conv2d(in_channels=in_dim, out_channels=in_dim, kernel_size=1)
        self.softmax = Softmax(dim=3)
        self.INF = INF
        self.gamma = nn.Parameter(torch.zeros(1))


    def forward(self, x):
        m_batchsize, _, height, width = x.size()
        proj_query = self.query_conv(x)
        proj_query_H = proj_query.permute(0,3,1,2).contiguous().view(m_batchsize*width,-1,height).permute(0, 2, 1)
        proj_query_W = proj_query.permute(0,2,1,3).contiguous().view(m_batchsize*height,-1,width).permute(0, 2, 1)
        proj_key = self.key_conv(x)
        proj_key_H = proj_key.permute(0,3,1,2).contiguous().view(m_batchsize*width,-1,height)
        proj_key_W = proj_key.permute(0,2,1,3).contiguous().view(m_batchsize*height,-1,width)
        proj_value = self.value_conv(x)
        proj_value_H = proj_value.permute(0,3,1,2).contiguous().view(m_batchsize*width,-1,height)
        proj_value_W = proj_value.permute(0,2,1,3).contiguous().view(m_batchsize*height,-1,width)
        energy_H = (torch.bmm(proj_query_H, proj_key_H)+self.INF(m_batchsize, height, width)).view(m_batchsize,width,height,height).permute(0,2,1,3)
        energy_W = torch.bmm(proj_query_W, proj_key_W).view(m_batchsize,height,width,width)
        concate = self.softmax(torch.cat([energy_H, energy_W], 3))

        att_H = concate[:,:,:,0:height].permute(0,2,1,3).contiguous().view(m_batchsize*width,height,height)

        att_W = concate[:,:,:,height:height+width].contiguous().view(m_batchsize*height,width,width)
        out_H = torch.bmm(proj_value_H, att_H.permute(0, 2, 1)).view(m_batchsize,width,-1,height).permute(0,2,3,1)
        out_W = torch.bmm(proj_value_W, att_W.permute(0, 2, 1)).view(m_batchsize,height,-1,width).permute(0,2,1,3)

        return self.gamma*(out_H + out_W) + x

语义分割中的nonlocal[3]-CCNet

原文：https://www.cnblogs.com/treeaxx/p/14432029.html