MobileNetV2: Inverted residuals and linear bottlenecks
论文链接：https://arxiv.org/pdf/1801.04381v3.pdf
PyTorch: https://github.com/shanglianlm0525/PyTorch-Networks
Mobilenet V2 Units
PyTorch代码：
import torch
import torch.nn as nn
import torchvision
from functools import reduce
def Conv3x3BNReLU(in_channels,out_channels,stride,groups):
return nn.Sequential(
nn.Conv2d(in_channels=in_channels, out_channels=out_channels, kernel_size=3, stride=stride, padding=1, groups=groups),
nn.BatchNorm2d(out_channels),
nn.ReLU6(inplace=True)
)
def Conv1x1BNReLU(in_channels,out_channels):
return nn.Sequential(
nn.Conv2d(in_channels=in_channels, out_channels=out_channels, kernel_size=1, stride=1),
nn.BatchNorm2d(out_channels),
nn.ReLU6(inplace=True)
)
def Conv1x1BN(in_channels,out_channels):
return nn.Sequential(
nn.Conv2d(in_channels=in_channels, out_channels=out_channels, kernel_size=1, stride=1),
nn.BatchNorm2d(out_channels)
)
class InvertedResidual(nn.Module):
def __init__(self, in_channels, out_channels, stride, expansion_factor=6):
super(InvertedResidual, self).__init__()
self.stride = stride
mid_channels = (in_channels * expansion_factor)
self.bottleneck = nn.Sequential(
Conv1x1BNReLU(in_channels, mid_channels),
Conv3x3BNReLU(mid_channels, mid_channels, stride,groups=mid_channels),
Conv1x1BN(mid_channels, out_channels)
)
if self.stride == 1:
self.shortcut = Conv1x1BN(in_channels, out_channels)
def forward(self, x):
out = self.bottleneck(x)
out = (out+self.shortcut(x)) if self.stride==1 else out
return out
class MobileNetV2(nn.Module):
def __init__(self, num_classes=1000):
super(MobileNetV2,self).__init__()
self.first_conv = Conv3x3BNReLU(3,32,2,groups=1)
self.layer1 = self.make_layer(in_channels=32, out_channels=16, stride=1, block_num=1)
self.layer2 = self.make_layer(in_channels=16, out_channels=24, stride=2, block_num=2)
self.layer3 = self.make_layer(in_channels=24, out_channels=32, stride=2, block_num=3)
self.layer4 = self.make_layer(in_channels=32, out_channels=64, stride=2, block_num=4)
self.layer5 = self.make_layer(in_channels=64, out_channels=96, stride=1, block_num=3)
self.layer6 = self.make_layer(in_channels=96, out_channels=160, stride=2, block_num=3)
self.layer7 = self.make_layer(in_channels=160, out_channels=320, stride=1, block_num=1)
self.last_conv = Conv1x1BNReLU(320,1280)
self.avgpool = nn.AvgPool2d(kernel_size=7,stride=1)
self.dropout = nn.Dropout(p=0.2)
self.linear = nn.Linear(in_features=1280,out_features=num_classes)
def make_layer(self, in_channels, out_channels, stride, block_num):
layers = []
layers.append(InvertedResidual(in_channels, out_channels, stride))
for i in range(1, block_num):
layers.append(InvertedResidual(out_channels,out_channels,1))
return nn.Sequential(*layers)
def init_params(self):
for m in self.modules():
if isinstance(m, nn.Conv2d):
nn.init.kaiming_normal_(m.weight)
nn.init.constant_(m.bias, 0)
elif isinstance(m, nn.Linear) or isinstance(m, nn.BatchNorm2d):
nn.init.constant_(m.weight, 1)
nn.init.constant_(m.bias, 0)
def forward(self, x):
x = self.first_conv(x)
x = self.layer1(x)
x = self.layer2(x)
x = self.layer3(x)
x = self.layer4(x)
x = self.layer5(x)
x = self.layer6(x)
x = self.layer7(x)
x = self.last_conv(x)
x = self.avgpool(x)
x = x.view(x.size(0),-1)
x = self.dropout(x)
out = self.linear(x)
return out
if __name__=='__main__':
model = MobileNetV2()
# model = torchvision.models.MobileNetV2()
print(model)
input = torch.randn(1, 3, 224, 224)
out = model(input)
print(out.shape)
