Non-local-net.py

import torch
from torch import nn
from torch.nn import functional as F
from mmcv.cnn import constant_init, kaiming_init

__all__ = ['NL_resnet50']

def last_zero_init(m):
    if isinstance(m, nn.Sequential):
        constant_init(m[-1], val=0)
        m[-1].inited = True
    else:
        constant_init(m, val=0)
        m.inited = True


class NLblock2d(nn.Module):

    def __init__(self, inplanes, planes, pool='att', fusions=['channel_add']):
        super(NLblock2d, self).__init__()
        assert pool in ['avg', 'att']
        assert all([f in ['channel_add', 'channel_mul'] for f in fusions])
        assert len(fusions) > 0, 'at least one fusion should be used'
        self.inplanes = inplanes
        self.planes = planes
        self.pool = pool
        self.fusions = fusions
        if 'att' in pool:
            self.conv_mask1 = nn.Conv2d(inplanes, inplanes, kernel_size=1, bias=False)
            self.conv_mask2 = nn.Conv2d(inplanes, inplanes, kernel_size=1, bias=False)
            self.conv_mask3 = nn.Conv2d(inplanes, inplanes, kernel_size=1, bias=False)
        if 'channel_add' in fusions:
            self.channel_add_conv = nn.Sequential(nn.Conv2d(self.inplanes, self.planes, kernel_size=1, bias=False))
         
        else:
            self.channel_add_conv = None

        if 'channel_mul' in fusions:
            self.channel_mul_conv = nn.Sequential(nn.Conv2d(self.inplanes, self.planes, kernel_size=1, bias=False))
             
        else:
            self.channel_mul_conv = None
        self.reset_parameters()

    def reset_parameters(self):
        if self.pool == 'att':
            kaiming_init(self.conv_mask1, mode='fan_in')
            self.conv_mask1.inited = True
            kaiming_init(self.conv_mask2, mode='fan_in')
            self.conv_mask2.inited = True
            kaiming_init(self.conv_mask3, mode='fan_in')
            self.conv_mask3.inited = True

        if self.channel_add_conv is not None:
            last_zero_init(self.channel_add_conv)
        if self.channel_mul_conv is not None:
            last_zero_init(self.channel_mul_conv)

    def spatial_pool(self, x):
        batch = x.size(0)
        if self.pool == 'att':
            input_x = x
            conv_mask1=self.conv_mask1(x)
            conv_mask2=self.conv_mask2(x)
            conv_mask3=self.conv_mask3(x)
            # [N, C, H * W]
            conv_mask1=conv_mask1.view(batch, self.inplanes, -1)
            conv_mask1=conv_mask1.permute(0, 2, 1)
            conv_mask2=conv_mask2.view(batch, self.inplanes, -1)
            conv_mask2=conv_mask2.permute(0, 2, 1)
            conv_mask3=conv_mask3.view(batch, self.inplanes, -1)
            # [N, 1, H * W, C]
#            conv_mask2 = conv_mask2.unsqueeze(1)
            # [N, 1, C, H * W]
#            conv_mask3 = conv_mask3.unsqueeze(1)
            # [N, 1, H * W, H * W]
            context = torch.matmul(conv_mask2, conv_mask3)
            # [N, 1, H * W, H * W]
            context = F.softmax(context, dim=-1)#softmax操作
            # [N, 1, H * W, C]
#            conv_mask1 = conv1_mask.unsqueeze(1)
            # [N, 1, H * W, C]
            context = torch.matmul(context, conv_mask1)
            # [N, C, H * W]
            context = context.permute(0, 2, 1).contiguous()
            context = context.view(batch, self.inplanes, *x.size()[2:])
        else:

            context = self.avg_pool(x)

        return context

    def forward(self, x):
        # [N, C, 1, 1]
        context = self.spatial_pool(x)

        if self.channel_mul_conv is not None:
            # [N, C, 1, 1]
            channel_mul_term = torch.sigmoid(self.channel_mul_conv(context))
            out = x * channel_mul_term
        else:
            out = x
        if self.channel_add_conv is not None:
            # [N, C, 1, 1]
            channel_add_term = self.channel_add_conv(context)
            out = out + channel_add_term

        return out


def conv3x3(in_planes, out_planes, stride=1):
    "3x3 convolution with padding"
    return nn.Conv2d(in_planes, out_planes, kernel_size=3, stride=stride,
                     padding=1, bias=False)


class BasicBlock(nn.Module):
    expansion = 1

    def __init__(self, inplanes, planes, stride=1, downsample=None):
        super(BasicBlock, self).__init__()
        self.conv1 = conv3x3(inplanes, planes, stride)
        self.bn1 = nn.BatchNorm2d(planes)
        self.relu = nn.ReLU(inplace=True)
        self.conv2 = conv3x3(planes, planes)
        self.bn2 = nn.BatchNorm2d(planes)
        self.downsample = downsample
        self.stride = stride

    def forward(self, x):
        residual = x

        out = self.conv1(x)
        out = self.bn1(out)
        out = self.relu(out)

        out = self.conv2(out)
        out = self.bn2(out)

        if self.downsample is not None:
            residual = self.downsample(x)

        out += residual
        out = self.relu(out)

        return out


class BottleNeck(nn.Module):
    expansion = 4

    def __init__(self, inplanes, planes, stride=1, downsample=None, nl=False):
        super(BottleNeck, self).__init__()
        self.conv1 = nn.Conv2d(inplanes, planes, kernel_size=1, bias=False)
        self.bn1 = nn.BatchNorm2d(planes)
        self.conv2 = nn.Conv2d(planes, planes, kernel_size=3, stride=stride,
                               padding=1, bias=False)
        self.bn2 = nn.BatchNorm2d(planes)
        self.conv3 = nn.Conv2d(planes, planes * 4, kernel_size=1, bias=False)
        self.bn3 = nn.BatchNorm2d(planes * 4)
        self.relu = nn.ReLU(inplace=True)
        self.downsample = downsample
        self.stride = stride

    def forward(self, x):
        residual = x

        out = self.conv1(x)
        out = self.bn1(out)
        out = self.relu(out)

        out = self.conv2(out)
        out = self.bn2(out)
        out = self.relu(out)

        out = self.conv3(out)
        out = self.bn3(out)

        if self.downsample is not None:
            residual = self.downsample(x)

        out += residual
        out = self.relu(out)

        return out

class ResNet(nn.Module):

    def __init__(self, block, block1, num_block, num_classes=100, nl=False):
        super().__init__()

        self.inplanes = 64

        self.conv1 = nn.Sequential(
            nn.Conv2d(3, 64, kernel_size=3, padding=1, bias=False),
            nn.BatchNorm2d(64),
            nn.ReLU(inplace=True))
        #we use a different inputsize than the original paper
        #so conv2_x's stride is 1
        self.conv2_x = self._make_layer(block, 64, num_block[0], 1)
        self.conv3_x = self._make_layer(block, 128, num_block[1], 2)
        self.conv4_x = self._make_layer1(block, block1, 256, num_block[2], 2, nl=nl)
        self.conv5_x = self._make_layer(block, 512, num_block[3], 2)
        self.avg_pool = nn.AdaptiveAvgPool2d((1, 1))
        self.fc = nn.Linear(512 * block.expansion, num_classes)

    def _make_layer(self, block, planes, blocks, stride=1):
        downsample = None
        if stride != 1 or self.inplanes != planes * block.expansion:
            downsample = nn.Sequential(
                nn.Conv2d(self.inplanes, planes * block.expansion,
                          kernel_size=1, stride=stride, bias=False),
                nn.BatchNorm2d(planes * block.expansion),
            )

        layers = []
        layers.append(block(self.inplanes, planes, stride, downsample))
        self.inplanes = planes * block.expansion
        for i in range(1, blocks):
            layers.append(block(self.inplanes, planes))
        
        return nn.Sequential(*layers)


    def _make_layer1(self, block, block1, planes, blocks, stride=1, nl=False):
        downsample = None
        if stride != 1 or self.inplanes != planes * block.expansion:
            downsample = nn.Sequential(
                nn.Conv2d(self.inplanes, planes * block.expansion,
                          kernel_size=1, stride=stride, bias=False),
                nn.BatchNorm2d(planes * block.expansion),
            )

        layers = []
        layers.append(block(self.inplanes, planes, stride, downsample))
        self.inplanes = planes * block.expansion
        for i in range(1, blocks-1):
            layers.append(block(self.inplanes, planes))
        layers.append(block1(self.inplanes, self.inplanes))
        layers.append(block(self.inplanes, planes))

        return nn.Sequential(*layers)


    def forward(self, x):
        output = self.conv1(x)
        output = self.conv2_x(output)
        output = self.conv3_x(output)
        output = self.conv4_x(output)
        output = self.conv5_x(output)
        output = self.avg_pool(output)
        output = output.view(output.size(0), -1)
        output = self.fc(output)

        return output 

def resnet18():
    """ return a ResNet 18 object
    """
    return ResNet(BasicBlock, [2, 2, 2, 2])

def resnet34():
    """ return a ResNet 34 object
    """
    return ResNet(BasicBlock, [3, 4, 6, 3])

def NL_resnet50():
    """ return a ResNet 50 object
    """
    return ResNet(BottleNeck, NLblock2d, [3, 4, 6, 3], nl=True)

def resnet101():
    """ return a ResNet 101 object
    """
    return ResNet(BottleNeck, [3, 4, 23, 3])

def resnet152():
    """ return a ResNet 152 object
    """
    return ResNet(BottleNeck, [3, 8, 36, 3])