pytorch中残差网络resnet的源码解读
pytorch的resnet模块在torchvision的models中。
里面可以选择的resnet类型有：
_all_列表的每一个resnet都提供了实现的函数：
def resnet18(pretrained=False, progress=True, **kwargs):
"""Constructs a ResNet-18 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
progress (bool): If True, displays a progress bar of the download to stderr
"""
return _resnet('resnet18', BasicBlock, [2, 2, 2, 2], pretrained, progress,
**kwargs)
def resnet34(pretrained=False, progress=True, **kwargs):
"""Constructs a ResNet-34 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
progress (bool): If True, displays a progress bar of the download to stderr
"""
return _resnet('resnet34', BasicBlock, [3, 4, 6, 3], pretrained, progress,
**kwargs)
def resnet50(pretrained=False, progress=True, **kwargs):
"""Constructs a ResNet-50 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
progress (bool): If True, displays a progress bar of the download to stderr
"""
return _resnet('resnet50', Bottleneck, [3, 4, 6, 3], pretrained, progress,
**kwargs)
def resnet101(pretrained=False, progress=True, **kwargs):
"""Constructs a ResNet-101 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
progress (bool): If True, displays a progress bar of the download to stderr
"""
return _resnet('resnet101', Bottleneck, [3, 4, 23, 3], pretrained, progress,
**kwargs)
def resnet152(pretrained=False, progress=True, **kwargs):
"""Constructs a ResNet-152 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
progress (bool): If True, displays a progress bar of the download to stderr
"""
return _resnet('resnet152', Bottleneck, [3, 8, 36, 3], pretrained, progress,
**kwargs)
def resnext50_32x4d(pretrained=False, progress=True, **kwargs):
"""Constructs a ResNeXt-50 32x4d model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
progress (bool): If True, displays a progress bar of the download to stderr
"""
kwargs['groups'] = 32
kwargs['width_per_group'] = 4
return _resnet('resnext50_32x4d', Bottleneck, [3, 4, 6, 3],
pretrained, progress, **kwargs)
def resnext101_32x8d(pretrained=False, progress=True, **kwargs):
"""Constructs a ResNeXt-101 32x8d model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
progress (bool): If True, displays a progress bar of the download to stderr
"""
kwargs['groups'] = 32
kwargs['width_per_group'] = 8
return _resnet('resnext101_32x8d', Bottleneck, [3, 4, 23, 3],
pretrained, progress, **kwargs)
一，简单介绍残差网络Resnet：
resnet是由很多以下的结构组成：
这种结构，当有1x1卷积核的时候，我们叫bottleneck，当没有1x1卷积核时，我们称其为BasicBlock。残差网络一般就是由这两个结构组成的。
残差网络的结构：（例resnet18）
(彩图resnet18的结构图中，虚曲线表示不同维度的连接，实曲线表示相同维度的连接)
从上图可以看到几个重点的关于resnet的特点：
1.resnet18都是由BasicBlock组成的，并且从表中也可以得知，50层(包括50层)以上的resnet才由Bottleneck组成。
2.所有类型的resnet卷积操作的通道数(无论是输入通道还是输出通道)都是64的倍数
3.所有类型的resnet的卷积核只有3x3和1x1两种
4.无论哪一种resnet，除了公共部分(conv1)外，都是由4大块组成(con2_x,con3_x,con4_x,con5_x,)，每一块的起始通道数都是64，128，256，512，这点非常重要。暂且称它为“基准 通道数”
了解这些有利于我们理解resnet的源码。
二，代码解析
1.两个重要结构：BasicBlock和Bottleneck
在pytorch的resnet模块在torchvision的models中。由于resnet中的卷积核不是1x1就是3x3，所以首先定义了这两个卷积核的操作：
def conv3x3(in_planes, out_planes, stride=1, groups=1, dilation=1):
"""3x3 convolution with padding"""
return nn.Conv2d(in_planes, out_planes, kernel_size=3, stride=stride,
padding=dilation, groups=groups, bias=False, dilation=dilation)
def conv1x1(in_planes, out_planes, stride=1):
"""1x1 convolution"""
return nn.Conv2d(in_planes, out_planes, kernel_size=1, stride=stride, bias=False)
作为重头戏的BasicBlock和Bottleneck都是以类的形式书写：
BasicBlock：
class BasicBlock(nn.Module):
expansion = 1 #expansion是BasicBlock和Bottleneck的核心区别之一
def __init__(self, inplanes, planes, stride=1, downsample=None, groups=1,
base_width=64, dilation=1, norm_layer=None):
super(BasicBlock, self).__init__()
if norm_layer is None:
norm_layer = nn.BatchNorm2d
if groups != 1 or base_width != 64:
raise ValueError('BasicBlock only supports groups=1 and base_width=64')
if dilation > 1:
raise NotImplementedError("Dilation > 1 not supported in BasicBlock")
# Both self.conv1 and self.downsample layers downsample the input when stride != 1
self.conv1 = conv3x3(inplanes, planes, stride)
self.bn1 = norm_layer(planes)
self.relu = nn.ReLU(inplace=True)
self.conv2 = conv3x3(planes, planes)
self.bn2 = norm_layer(planes)
self.downsample = downsample
self.stride = stride
def forward(self, x):
identity = 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:
identity = self.downsample(x)
out += identity
out = self.relu(out)
return out看到代码 self.downsample = downsample，在默认情况downsample=None，表示不做downsample，但有一个情况需要做，就是一个 BasicBlock的分支x要与output相加时，若x和output的通道数不一样，则要做一个downsample，剧透一下，在resnet里的downsample就是用一个1x1的卷积核处理，变成想要的通道数。为什么要这样做？因为最后要x要和output相加啊， 通道不同相加不了。所以downsample是专门用来改变x的通道数的。
接下来分析BasicBlock处理后的图像的维度是如何变化的：
我们看到，BasicBlock虽然是经过两个3x3的卷积，但是前一个是设置了步长的(剧透给你，设置的步长是2)，后一个则是没有设置步长的，这意味着用的是默认步长(剧透给你，默认步长是1)，接下来也不想剧透了，直接给看con3x3的定义吧：
可以看到padding在没指定的情况下，默认也是1.根据公式：
W为特征图的长度(或宽度)，F为卷积核的长度(或宽度)，P为padding，S为步长。所以卷积后的特征图的尺寸就跟步长很有关系，我们知道卷积核为3x3，所以F=3，也知道P=1.所以 当S=1时，W是不变的。当S=2时，W会减少两倍。下面的Bottleneck也是这个原理。就不多写了。
BasicBlock即：
接下来是
Bottleneck：
class Bottleneck(nn.Module):
expansion = 4 #expansion是BasicBlock和Bottleneck的核心区别之一
def __init__(self, inplanes, planes, stride=1, downsample=None, groups=1,
base_width=64, dilation=1, norm_layer=None):
super(Bottleneck, self).__init__()
if norm_layer is None:
norm_layer = nn.BatchNorm2d
width = int(planes * (base_width / 64.)) * groups
# Both self.conv2 and self.downsample layers downsample the input when stride != 1
self.conv1 = conv1x1(inplanes, width)
self.bn1 = norm_layer(width)
self.conv2 = conv3x3(width, width, stride, groups, dilation)
self.bn2 = norm_layer(width)
self.conv3 = conv1x1(width, planes * self.expansion)
self.bn3 = norm_layer(planes * self.expansion)
self.relu = nn.ReLU(inplace=True)
self.downsample = downsample
self.stride = stride
def forward(self, x):
identity = 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:
identity = self.downsample(x)
out += identity
out = self.relu(out)
return outBottleneck即：
BasicBlock和Bottleneck的两点核心区别：
1.BasicBlock的卷积核都是2个3x3，Bottleneck则是一个1x1，3x3，1x1共三个卷积核组成。
2.BasicBlock的expansion为1，即输入和输出的通道数是一致的。而Bottleneck的expansion为4，即输出通道数是输入通道数的4倍。
了解这些有利于代码 的理解。
关于downsample：
不管是BasicBlock还是Bottleneck，最后都会做一个判断是否需要给x做downsample，因为必须要把x的通道数变成与主枝的输出的通道一致，才能相加。
2.resnet的本体(ResNet类)
源码中的ResNet类可以根据输入参数的不同，变成resnet18，34，50，101等。
这个ResNet非常重要，由于考虑篇幅问题，后面再细讲。
3.调用pytorch源码使用resnet：
from torchvision.models.resnet import resnet50, Bottleneck
resnet = resnet50(pretrained=True)短短两行代码就可以调用resnet了。接下来我们顺着源码“顺藤摸瓜”地看一下代码的执行流程：
从resnet50(pretrained=True)可以看出，调用的是resnet50网络，然后通过在torchvision.models.resnet模块中可以找到resnet50()的定义：
def resnet50(pretrained=False, progress=True, **kwargs):
"""Constructs a ResNet-50 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
progress (bool): If True, displays a progress bar of the download to stderr
"""
return _resnet('resnet50', Bottleneck, [3, 4, 6, 3], pretrained, progress,
**kwargs)
可以看到，resnet50()仅仅做了一件事，就是调用了配备了特定参数的_resnet()方法。
即_resnet('resnet50', Bottleneck, [3, 4, 6, 3], pretrained, progress, **kwargs)
第一个参数不用说了，而我们看到 Bottleneck参数，就可以联想到_resnet的第二个参数是resnet网络的组成部分，即不是BasicBlock就是Bottleneck。而[3，4，6，3]具体是什么意思我们要接着看_resnet()的定义：
def _resnet(arch, block, layers, pretrained, progress, **kwargs):
model = ResNet(block, layers, **kwargs)
if pretrained:
state_dict = load_state_dict_from_url(model_urls[arch],
progress=progress)
model.load_state_dict(state_dict)
return model我们可以看到在_resnet()的[3，4，6，3]对应的位置，是显示layers的，且把其作为参数传给了残差网络的本体类Resnet。
接着判断是否需要预训练(pretrained是否为True)，为True则加载权重后返回模型，为False就直接返回模型。
而_resnet()中block参数对应的位置就是BasicBlock或Bottleneck。即block就是表示BasicBlock或Bottleneck。
所以一句话概括_resnet()的作用就是先调用Resnet类生成一个Resnet的壳子，若需要预训练则加载权重，不需要就直接返回Resnet的壳子。
以resnet50为例，_resnet()的arch参数填入的是'resnet50'，然后会根据这个参数到model_urls字典中找寻相对应的权重资源下载。
该字典如下：
===================
接下来就仔细看看Resnet类的代码了：
ResNet：
class ResNet(nn.Module):
def __init__(self, block, layers, num_classes=1000, zero_init_residual=False,
groups=1, width_per_group=64, replace_stride_with_dilation=None,
norm_layer=None):
super(ResNet, self).__init__()
if norm_layer is None:
norm_layer = nn.BatchNorm2d
self._norm_layer = norm_layer
self.inplanes = 64
self.dilation = 1
if replace_stride_with_dilation is None:
# each element in the tuple indicates if we should replace
# the 2x2 stride with a dilated convolution instead
replace_stride_with_dilation = [False, False, False]
if len(replace_stride_with_dilation) != 3:
raise ValueError("replace_stride_with_dilation should be None "
"or a 3-element tuple, got {}".format(replace_stride_with_dilation))
self.groups = groups
self.base_width = width_per_group
self.conv1 = nn.Conv2d(3, self.inplanes, kernel_size=7, stride=2, padding=3,
bias=False)
self.bn1 = norm_layer(self.inplanes)
self.relu = nn.ReLU(inplace=True)
self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1)
self.layer1 = self._make_layer(block, 64, layers[0])
self.layer2 = self._make_layer(block, 128, layers[1], stride=2,
dilate=replace_stride_with_dilation[0])
self.layer3 = self._make_layer(block, 256, layers[2], stride=2,
dilate=replace_stride_with_dilation[1])
self.layer4 = self._make_layer(block, 512, layers[3], stride=2,
dilate=replace_stride_with_dilation[2])
self.avgpool = nn.AdaptiveAvgPool2d((1, 1))
self.fc = nn.Linear(512 * block.expansion, num_classes)
for m in self.modules():
if isinstance(m, nn.Conv2d):
nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu')
elif isinstance(m, (nn.BatchNorm2d, nn.GroupNorm)):
nn.init.constant_(m.weight, 1)
nn.init.constant_(m.bias, 0)
# Zero-initialize the last BN in each residual branch,
# so that the residual branch starts with zeros, and each residual block behaves like an identity.
# This improves the model by 0.2~0.3% according to https://arxiv.org/abs/1706.02677
if zero_init_residual:
for m in self.modules():
if isinstance(m, Bottleneck):
nn.init.constant_(m.bn3.weight, 0)
elif isinstance(m, BasicBlock):
nn.init.constant_(m.bn2.weight, 0)
def _make_layer(self, block, planes, blocks, stride=1, dilate=False):
norm_layer = self._norm_layer
downsample = None
previous_dilation = self.dilation
if dilate:
self.dilation *= stride
stride = 1
if stride != 1 or self.inplanes != planes * block.expansion:
downsample = nn.Sequential(
conv1x1(self.inplanes, planes * block.expansion, stride),
norm_layer(planes * block.expansion),
)
layers = []
layers.append(block(self.inplanes, planes, stride, downsample, self.groups,
self.base_width, previous_dilation, norm_layer))
self.inplanes = planes * block.expansion
for _ in range(1, blocks):
layers.append(block(self.inplanes, planes, groups=self.groups,
base_width=self.base_width, dilation=self.dilation,
norm_layer=norm_layer))
return nn.Sequential(*layers)
def forward(self, x):
x = self.conv1(x)
x = self.bn1(x)
x = self.relu(x)
x = self.maxpool(x)
x = self.layer1(x)
x = self.layer2(x)
x = self.layer3(x)
x = self.layer4(x)
x = self.avgpool(x)
x = x.reshape(x.size(0), -1)
x = self.fc(x)
return x一眼看下去不知道该从何入手，但看pytorch的网络，要想知道它的执行顺序是怎么样的，看它的forward方法即可。
从ResNet的forward代码来看，它是先经过conv1()，bn，relu和maxpool()。从之前的resnet表格得知，这几层无论是resnet18，resnet34，resnet50，resnet101等等的resnet一开始都必须经过这几层。这是静态的。
然后进入四层layer()才是动态以区别具体是resnet18，resnet34，resnet50，resnet101等等中的哪一个。
从动态代码来看，所有的resnet都有4个动态层。追寻layer1，2，3，4究竟是如何定义的我们可以看到：
其实所谓的layer1，2，3，4都是由不同参数的_make_layer()方法得到的。看_make_layer()的参数，发现了layers[0~3]就是上面输入的[3，4，6，3]，即layers[0]是3，layers[1]是4，layers[2]是6，layers[3]是3。我们继续追寻_make_layer()的定义看看这些数字表示什么意思：
def _make_layer(self, block, planes, blocks, stride=1, dilate=False):
norm_layer = self._norm_layer
downsample = None
previous_dilation = self.dilation
if dilate:
self.dilation *= stride
stride = 1
if stride != 1 or self.inplanes != planes * block.expansion:
downsample = nn.Sequential(
conv1x1(self.inplanes, planes * block.expansion, stride),
norm_layer(planes * block.expansion),
)
layers = []
layers.append(block(self.inplanes, planes, stride, downsample, self.groups,
self.base_width, previous_dilation, norm_layer))
self.inplanes = planes * block.expansion
for _ in range(1, blocks):
layers.append(block(self.inplanes, planes, groups=self.groups,
base_width=self.base_width, dilation=self.dilation,
norm_layer=norm_layer))
return nn.Sequential(*layers)(注意：_make_layer()中的planes参数是“基准通道数”，不是输出通道数！！！不是输出通道数！！！不是输出通道数！！！)
我们定位到_make_layer()的第三个(不算上self)参数blocks，在_make_layer()中用到blocks的地方是：
layers = []
layers.append(block(self.inplanes, planes, stride, downsample, self.groups,
self.base_width, previous_dilation, norm_layer))
self.inplanes = planes * block.expansion
for _ in range(1, blocks):
layers.append(block(self.inplanes, planes, groups=self.groups,
base_width=self.base_width, dilation=self.dilation,
norm_layer=norm_layer))可以看到其实blocks这个整数就是表示生成的block的数目。由于在resnet50填入的block是Bottleneck，所以blocks表示Bottleneck的数目。因此[3，4，6，3]表示按次序生成3个Bottleneck，4个Bottleneck，6个Bottleneck，3个Bottleneck。这个表格中resnet50的结构是一致的。所以layers[0]是3个Bottleneck，layers[1]是4个Bottleneck，layers[2]是6个Bottleneck，layers[3]是3个Bottleneck。
===========================
综合来说：
==========================
接下来是resnet一些点的分析：
1.首先是输出维度：
我们看到output size那里，维度是依次下降两倍的，56-28-14-7，但是我们会不会有疑问，经过那么多个卷积核为什么才下降两倍呢？那我们来看看是怎么实现的，这个实现是与ResNet类中的_make_layer()方法密切相关：
看到_make_layer()源码，有没有怀疑过为什么都是建立block，但是要分开两个步骤建呢，是因为上面那个是设置了步长的，步长为2，下面那个是用默认步长的，步长为1。当步长为2时，结合卷积核分析，会减低2倍特征图尺寸。步长为1时，则不变。
总结一些小知识：
1.无论哪种resnet，都有4个layer，进入layer之前，输入图片就已经被缩小了4倍了(一个卷积和一个最大池化操作各1/2)。除了第一个layer不会缩小图片外，其余三个layer都会缩小一半图片。
==========================
最后附上完整的torchvision.models.resnet：
import torch.nn as nn
from .utils import load_state_dict_from_url
__all__ = ['ResNet', 'resnet18', 'resnet34', 'resnet50', 'resnet101',
'resnet152', 'resnext50_32x4d', 'resnext101_32x8d']
model_urls = {
'resnet18': 'https://download.pytorch.org/models/resnet18-5c106cde.pth',
'resnet34': 'https://download.pytorch.org/models/resnet34-333f7ec4.pth',
'resnet50': 'https://download.pytorch.org/models/resnet50-19c8e357.pth',
'resnet101': 'https://download.pytorch.org/models/resnet101-5d3b4d8f.pth',
'resnet152': 'https://download.pytorch.org/models/resnet152-b121ed2d.pth',
'resnext50_32x4d': 'https://download.pytorch.org/models/resnext50_32x4d-7cdf4587.pth',
'resnext101_32x8d': 'https://download.pytorch.org/models/resnext101_32x8d-8ba56ff5.pth',
}
def conv3x3(in_planes, out_planes, stride=1, groups=1, dilation=1):
"""3x3 convolution with padding"""
return nn.Conv2d(in_planes, out_planes, kernel_size=3, stride=stride,
padding=dilation, groups=groups, bias=False, dilation=dilation)
def conv1x1(in_planes, out_planes, stride=1):
"""1x1 convolution"""
return nn.Conv2d(in_planes, out_planes, kernel_size=1, stride=stride, bias=False)
class BasicBlock(nn.Module):
expansion = 1
def __init__(self, inplanes, planes, stride=1, downsample=None, groups=1,
base_width=64, dilation=1, norm_layer=None):
super(BasicBlock, self).__init__()
if norm_layer is None:
norm_layer = nn.BatchNorm2d
if groups != 1 or base_width != 64:
raise ValueError('BasicBlock only supports groups=1 and base_width=64')
if dilation > 1:
raise NotImplementedError("Dilation > 1 not supported in BasicBlock")
# Both self.conv1 and self.downsample layers downsample the input when stride != 1
self.conv1 = conv3x3(inplanes, planes, stride)
self.bn1 = norm_layer(planes)
self.relu = nn.ReLU(inplace=True)
self.conv2 = conv3x3(planes, planes)
self.bn2 = norm_layer(planes)
self.downsample = downsample
self.stride = stride
def forward(self, x):
identity = 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:
identity = self.downsample(x)
out += identity
out = self.relu(out)
return out
class Bottleneck(nn.Module):
expansion = 4
def __init__(self, inplanes, planes, stride=1, downsample=None, groups=1,
base_width=64, dilation=1, norm_layer=None):
super(Bottleneck, self).__init__()
if norm_layer is None:
norm_layer = nn.BatchNorm2d
width = int(planes * (base_width / 64.)) * groups
# Both self.conv2 and self.downsample layers downsample the input when stride != 1
self.conv1 = conv1x1(inplanes, width)
self.bn1 = norm_layer(width)
self.conv2 = conv3x3(width, width, stride, groups, dilation)
self.bn2 = norm_layer(width)
self.conv3 = conv1x1(width, planes * self.expansion)
self.bn3 = norm_layer(planes * self.expansion)
self.relu = nn.ReLU(inplace=True)
self.downsample = downsample
self.stride = stride
def forward(self, x):
identity = 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:
identity = self.downsample(x)
out += identity
out = self.relu(out)
return out
class ResNet(nn.Module):
def __init__(self, block, layers, num_classes=1000, zero_init_residual=False,
groups=1, width_per_group=64, replace_stride_with_dilation=None,
norm_layer=None):
super(ResNet, self).__init__()
if norm_layer is None:
norm_layer = nn.BatchNorm2d
self._norm_layer = norm_layer
self.inplanes = 64
self.dilation = 1
if replace_stride_with_dilation is None:
# each element in the tuple indicates if we should replace
# the 2x2 stride with a dilated convolution instead
replace_stride_with_dilation = [False, False, False]
if len(replace_stride_with_dilation) != 3:
raise ValueError("replace_stride_with_dilation should be None "
"or a 3-element tuple, got {}".format(replace_stride_with_dilation))
self.groups = groups
self.base_width = width_per_group
self.conv1 = nn.Conv2d(3, self.inplanes, kernel_size=7, stride=2, padding=3,
bias=False)
self.bn1 = norm_layer(self.inplanes)
self.relu = nn.ReLU(inplace=True)
self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1)
self.layer1 = self._make_layer(block, 64, layers[0])
self.layer2 = self._make_layer(block, 128, layers[1], stride=2,
dilate=replace_stride_with_dilation[0])
self.layer3 = self._make_layer(block, 256, layers[2], stride=2,
dilate=replace_stride_with_dilation[1])
self.layer4 = self._make_layer(block, 512, layers[3], stride=2,
dilate=replace_stride_with_dilation[2])
self.avgpool = nn.AdaptiveAvgPool2d((1, 1))
self.fc = nn.Linear(512 * block.expansion, num_classes)
for m in self.modules():
if isinstance(m, nn.Conv2d):
nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu')
elif isinstance(m, (nn.BatchNorm2d, nn.GroupNorm)):
nn.init.constant_(m.weight, 1)
nn.init.constant_(m.bias, 0)
# Zero-initialize the last BN in each residual branch,
# so that the residual branch starts with zeros, and each residual block behaves like an identity.
# This improves the model by 0.2~0.3% according to https://arxiv.org/abs/1706.02677
if zero_init_residual:
for m in self.modules():
if isinstance(m, Bottleneck):
nn.init.constant_(m.bn3.weight, 0)
elif isinstance(m, BasicBlock):
nn.init.constant_(m.bn2.weight, 0)
def _make_layer(self, block, planes, blocks, stride=1, dilate=False):
norm_layer = self._norm_layer
downsample = None
previous_dilation = self.dilation
if dilate:
self.dilation *= stride
stride = 1
if stride != 1 or self.inplanes != planes * block.expansion:
downsample = nn.Sequential(
conv1x1(self.inplanes, planes * block.expansion, stride),
norm_layer(planes * block.expansion),
)
layers = []
layers.append(block(self.inplanes, planes, stride, downsample, self.groups,
self.base_width, previous_dilation, norm_layer))
self.inplanes = planes * block.expansion
for _ in range(1, blocks):
layers.append(block(self.inplanes, planes, groups=self.groups,
base_width=self.base_width, dilation=self.dilation,
norm_layer=norm_layer))
return nn.Sequential(*layers)
def forward(self, x):
x = self.conv1(x)
x = self.bn1(x)
x = self.relu(x)
x = self.maxpool(x)
x = self.layer1(x)
x = self.layer2(x)
x = self.layer3(x)
x = self.layer4(x)
x = self.avgpool(x)
x = x.reshape(x.size(0), -1)
x = self.fc(x)
return x
def _resnet(arch, block, layers, pretrained, progress, **kwargs):
model = ResNet(block, layers, **kwargs)
if pretrained:
state_dict = load_state_dict_from_url(model_urls[arch],
progress=progress)
model.load_state_dict(state_dict)
return model
def resnet18(pretrained=False, progress=True, **kwargs):
"""Constructs a ResNet-18 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
progress (bool): If True, displays a progress bar of the download to stderr
"""
return _resnet('resnet18', BasicBlock, [2, 2, 2, 2], pretrained, progress,
**kwargs)
def resnet34(pretrained=False, progress=True, **kwargs):
"""Constructs a ResNet-34 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
progress (bool): If True, displays a progress bar of the download to stderr
"""
return _resnet('resnet34', BasicBlock, [3, 4, 6, 3], pretrained, progress,
**kwargs)
def resnet50(pretrained=False, progress=True, **kwargs):
"""Constructs a ResNet-50 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
progress (bool): If True, displays a progress bar of the download to stderr
"""
return _resnet('resnet50', Bottleneck, [3, 4, 6, 3], pretrained, progress,
**kwargs)
def resnet101(pretrained=False, progress=True, **kwargs):
"""Constructs a ResNet-101 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
progress (bool): If True, displays a progress bar of the download to stderr
"""
return _resnet('resnet101', Bottleneck, [3, 4, 23, 3], pretrained, progress,
**kwargs)
def resnet152(pretrained=False, progress=True, **kwargs):
"""Constructs a ResNet-152 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
progress (bool): If True, displays a progress bar of the download to stderr
"""
return _resnet('resnet152', Bottleneck, [3, 8, 36, 3], pretrained, progress,
**kwargs)
def resnext50_32x4d(pretrained=False, progress=True, **kwargs):
"""Constructs a ResNeXt-50 32x4d model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
progress (bool): If True, displays a progress bar of the download to stderr
"""
kwargs['groups'] = 32
kwargs['width_per_group'] = 4
return _resnet('resnext50_32x4d', Bottleneck, [3, 4, 6, 3],
pretrained, progress, **kwargs)
def resnext101_32x8d(pretrained=False, progress=True, **kwargs):
"""Constructs a ResNeXt-101 32x8d model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
progress (bool): If True, displays a progress bar of the download to stderr
"""
kwargs['groups'] = 32
kwargs['width_per_group'] = 8
return _resnet('resnext101_32x8d', Bottleneck, [3, 4, 23, 3],
pretrained, progress, **kwargs)