蚌埠大建设及棚户区改造官方网站,wordpress网站使用,家居企业网站建设新闻,品牌建设 深度学习网络模型 MobileNet系列MobileNet V1、MobileNet V2、MobileNet V3网络详解以及pytorch代码复现 1、DW卷积与普通卷积计算量对比DW与PW计算量普通卷积计算量计算量对比 2、MobileNet V1MobileNet V1网络结构MobileNet V1网络结构代码 3、MobileNet V2倒残差结构模块倒残… 深度学习网络模型 MobileNet系列MobileNet V1、MobileNet V2、MobileNet V3网络详解以及pytorch代码复现 1、DW卷积与普通卷积计算量对比DW与PW计算量普通卷积计算量计算量对比 2、MobileNet V1MobileNet V1网络结构MobileNet V1网络结构代码 3、MobileNet V2倒残差结构模块倒残差模块代码MobileNet V2详细网络结构MobileNet V2网络结构代码 4、MobileNet V3创新点MobileNet V3详细网络结构注意力机制SE模块代码InvertedResidual模块代码整体代码 pytorch代码复现MobileNet V1~V2项目目录 1、DW卷积与普通卷积计算量对比 DW与PW计算量 普通卷积计算量 计算量对比 因此理论上普通卷积是DWPW卷积的8到9倍 2、MobileNet V1
MobileNet V1网络结构 MobileNet V1网络结构代码
import torch.nn as nn
import torchclass MobileNetV1(nn.Module):def __init__(self, ch_in, n_classes):super(MobileNetV1, self).__init__()# 定义普通卷积、BN、激活模块def conv_bn(inp, oup, stride):return nn.Sequential(nn.Conv2d(inp, oup, 3, stride, 1, biasFalse),nn.BatchNorm2d(oup),nn.ReLU(inplaceTrue))# 定义DW、PW卷积模块def conv_dw(inp, oup, stride):return nn.Sequential(# dwnn.Conv2d(inp, inp, 3, stride, 1, groupsinp, biasFalse), # DW卷积的卷积核输入与输出的数量一致且等于分组数nn.BatchNorm2d(inp),nn.ReLU(inplaceTrue),# pwnn.Conv2d(inp, oup, 1, 1, 0, biasFalse),nn.BatchNorm2d(oup),nn.ReLU(inplaceTrue),)self.model nn.Sequential(conv_bn(ch_in, 32, 2),conv_dw(32, 64, 1),conv_dw(64, 128, 2),conv_dw(128, 128, 1),conv_dw(128, 256, 2),conv_dw(256, 256, 1),conv_dw(256, 512, 2),conv_dw(512, 512, 1),conv_dw(512, 512, 1),conv_dw(512, 512, 1),conv_dw(512, 512, 1),conv_dw(512, 512, 1),conv_dw(512, 1024, 2),conv_dw(1024, 1024, 1),nn.AdaptiveAvgPool2d(1))self.fc nn.Linear(1024, n_classes)def forward(self, x):x self.model(x)x x.view(-1, 1024)x self.fc(x)return xif __name____main__:# model checkmodel MobileNetV1(ch_in3, n_classes5)print(model)random_datatorch.rand([1,3,224,224])result model(random_data)print(result)
3、MobileNet V2 倒残差结构模块 Residual blok与Inverted residual block对比 Residual blok先采用1 x 1的卷积核来对特征矩阵进行压缩减少输入特征矩阵的channel再通过3 x 3的卷积核进行特征处理再采用1 x 1的卷积核来扩充channel维度形成了两头大中间小的瓶颈结构。并且3 x 3的卷积后面采用Relu激活函数。Inverted residual block先采用1 x 1的卷积核进行升高channel维度的操作通过卷积核大小为3 x 3的DW模块进行卷积再通过1 x 1的卷积进行降低channel维度的处理形成两头小中间大的结构。并且3 x 3的卷积后面采用Relu6激活函数。Relu6激活函数倒残差结构详细示意图 倒残差模块代码
# 定义普通卷积、BN结构
class ConvBNReLU(nn.Sequential):def __init__(self, in_channel, out_channel, kernel_size3, stride1, groups1):padding (kernel_size - 1) // 2 # padding的设置根据kernel_size来定如果kernel_size为3则padding设置为1如果kernel_size为1为padding为0super(ConvBNReLU, self).__init__(# 在pytorch中如果设置的 group1的话就为普通卷积如果设置的值为输入特征矩阵的深度的话即in_channel则为深度卷积deptwise conv并且Dw卷积的输出特征矩阵的深度等于输入特征矩阵的深度nn.Conv2d(in_channel, out_channel, kernel_size, stride, padding, groupsgroups, biasFalse), # groups1,表示普通的卷积因为接下来要使用的是BN层此处的偏置不起任何作用所以设置为1nn.BatchNorm2d(out_channel),nn.ReLU6(inplaceTrue) # 此处使用的是Relu6激活函数)
# 定义mobile网络基本结构--即到残差结构
class InvertedResidual(nn.Module):def __init__(self, in_channel, out_channel, stride, expand_ratio):super(InvertedResidual, self).__init__()hidden_channel in_channel * expand_ratioself.use_shortcut stride 1 and in_channel out_channel # stride 1 and in_channel out_channel保证输入矩阵与输出矩阵的shape一致且通道数也一致这样才可以进行shurtcutlayers []if expand_ratio ! 1: # 表示如果扩展因子不为1时则使用1x1的卷积层即对输入特征矩阵的深度进行扩充# 1x1 pointwise convlayers.append(ConvBNReLU(in_channel, hidden_channel, kernel_size1))layers.extend([# 3x3 depthwise conv# 在pytorch中如果设置的 group1的话就为普通卷积如果设置的值为输入特征矩阵的深度的话即in_channel则为深度卷积deptwise conv并且Dw卷积的输出特征矩阵的深度等于输入特征矩阵的深度ConvBNReLU(hidden_channel, hidden_channel, stridestride, groupshidden_channel),# 1x1 pointwise conv(linear) 因为其后跟随的是线性激活函数即yx所以其后面不在跟随激活函数nn.Conv2d(hidden_channel, out_channel, kernel_size1, biasFalse),nn.BatchNorm2d(out_channel),])self.conv nn.Sequential(*layers)def forward(self, x):if self.use_shortcut:return x self.conv(x)else:return self.conv(x)
MobileNet V2详细网络结构 MobileNet V2网络结构代码
from torch import nn
import torchdef _make_divisible(ch, divisor8, min_chNone):将输入的通道数(ch)调整到divisor的整数倍方便硬件加速This function is taken from the original tf repo.It ensures that all layers have a channel number that is divisible by 8It can be seen here:https://github.com/tensorflow/models/blob/master/research/slim/nets/mobilenet/mobilenet.pyif min_ch is None:min_ch divisornew_ch max(min_ch, int(ch divisor / 2) // divisor * divisor)# Make sure that round down does not go down by more than 10%.if new_ch 0.9 * ch:new_ch divisorreturn new_ch# 定义普通卷积、BN结构
class ConvBNReLU(nn.Sequential):def __init__(self, in_channel, out_channel, kernel_size3, stride1, groups1):padding (kernel_size - 1) // 2 # padding的设置根据kernel_size来定如果kernel_size为3则padding设置为1如果kernel_size为1为padding为0super(ConvBNReLU, self).__init__(# 在pytorch中如果设置的 group1的话就为普通卷积如果设置的值为输入特征矩阵的深度的话即in_channel则为深度卷积deptwise conv并且Dw卷积的输出特征矩阵的深度等于输入特征矩阵的深度nn.Conv2d(in_channel, out_channel, kernel_size, stride, padding, groupsgroups, biasFalse), # groups1,表示普通的卷积因为接下来要使用的是BN层此处的偏置不起任何作用所以设置为1nn.BatchNorm2d(out_channel),nn.ReLU6(inplaceTrue) # 此处使用的是Relu6激活函数)# 定义mobile网络基本结构--即到残差结构
class InvertedResidual(nn.Module):def __init__(self, in_channel, out_channel, stride, expand_ratio):super(InvertedResidual, self).__init__()hidden_channel in_channel * expand_ratioself.use_shortcut stride 1 and in_channel out_channel # stride 1 and in_channel out_channel保证输入矩阵与输出矩阵的shape一致且通道数也一致这样才可以进行shurtcutlayers []if expand_ratio ! 1: # 表示如果扩展因子不为1时则使用1x1的卷积层即对输入特征矩阵的深度进行扩充# 1x1 pointwise convlayers.append(ConvBNReLU(in_channel, hidden_channel, kernel_size1))layers.extend([# 3x3 depthwise conv# 在pytorch中如果设置的 group1的话就为普通卷积如果设置的值为输入特征矩阵的深度的话即in_channel则为深度卷积deptwise conv并且Dw卷积的输出特征矩阵的深度等于输入特征矩阵的深度ConvBNReLU(hidden_channel, hidden_channel, stridestride, groupshidden_channel),# 1x1 pointwise conv(linear) 因为其后跟随的是线性激活函数即yx所以其后面不在跟随激活函数nn.Conv2d(hidden_channel, out_channel, kernel_size1, biasFalse),nn.BatchNorm2d(out_channel),])self.conv nn.Sequential(*layers)def forward(self, x):if self.use_shortcut:return x self.conv(x)else:return self.conv(x)# 定义mobileNetV2网络
class MobileNetV2(nn.Module):def __init__(self, num_classes1000, alpha1.0, round_nearest8):super(MobileNetV2, self).__init__()block InvertedResidualinput_channel _make_divisible(32 * alpha, round_nearest) # 将卷积核的个数调整为8的整数倍last_channel _make_divisible(1280 * alpha, round_nearest)inverted_residual_setting [# t, c, n, s[1, 16, 1, 1],[6, 24, 2, 2],[6, 32, 3, 2],[6, 64, 4, 2],[6, 96, 3, 1],[6, 160, 3, 2],[6, 320, 1, 1],]features []# conv1 layerfeatures.append(ConvBNReLU(3, input_channel, stride2)) # 添加第一层普通卷积层# building inverted residual residual blockesfor t, c, n, s in inverted_residual_setting:output_channel _make_divisible(c * alpha, round_nearest) # 根据alpha因子调整卷积核的个数for i in range(n): # 循环添加倒残差模块stride s if i 0 else 1 # s表示的是倒残差模块结构中第一层卷积对应的步距剩余层都是1features.append(block(input_channel, output_channel, stride, expand_ratiot)) # 添加一系列倒残差结构input_channel output_channel# building last several layersfeatures.append(ConvBNReLU(input_channel, last_channel, 1)) # 构建最后一层卷积层# combine feature layersself.features nn.Sequential(*features)# building classifierself.avgpool nn.AdaptiveAvgPool2d((1, 1)) # 采用自适应平均采样层self.classifier nn.Sequential(nn.Dropout(0.2),nn.Linear(last_channel, num_classes))# weight initialization 初始化全只能怪for m in self.modules():if isinstance(m, nn.Conv2d):nn.init.kaiming_normal_(m.weight, modefan_out)if m.bias is not None:nn.init.zeros_(m.bias)elif isinstance(m, nn.BatchNorm2d):nn.init.ones_(m.weight)nn.init.zeros_(m.bias)elif isinstance(m, nn.Linear):nn.init.normal_(m.weight, 0, 0.01) # 初始化为正态分布的函数均值为0方差为0.01nn.init.zeros_(m.bias)def forward(self, x):x self.features(x)x self.avgpool(x)x torch.flatten(x, 1)x self.classifier(x)return xif __name__ __main__:divisible _make_divisible(1)print(divisible)4、MobileNet V3 创新点 加入了注意力机制SE模块使用的新的激活函数 激活函数 MobileNet V3详细网络结构 # 定义block的配置类
class InvertedResidualConfig:def __init__(self,input_c: int, # block模块中的第一个1x1卷积层的输入channel数kernel: int, # depthwise卷积的卷积核大小expanded_c: int, # block模块中的第一个1x1卷积层的输出channel数out_c: int, # 经过block模块中第二个1x1卷积层处理过后得到的channel数use_se: bool, # 是否使用注意力机制模块activation: str, # 激活方式stride: int, # 步长width_multi: float): # width_multi调节每个卷积层所使用channel的倍率因子self.input_c self.adjust_channels(input_c, width_multi)self.kernel kernelself.expanded_c self.adjust_channels(expanded_c, width_multi)self.out_c self.adjust_channels(out_c, width_multi)self.use_se use_seself.use_hs activation HS # whether using h-swish activationself.stride stridestaticmethoddef adjust_channels(channels: int, width_multi: float):return _make_divisible(channels * width_multi, 8) 注意力机制SE模块代码
# 注意力机制模块SE模块即两个全连接层 该模块的基本流程是先进行自适应平均池化(1x1)———1x1的卷积层———relu激活层———1x1的卷积池化———hardsigmoid()激活函数激活
class SqueezeExcitation(nn.Module):def __init__(self, input_c: int, squeeze_factor: int 4):super(SqueezeExcitation, self).__init__()squeeze_c _make_divisible(input_c // squeeze_factor, 8) # 获得距离该数最近的8的整数倍的数字self.fc1 nn.Conv2d(input_c, squeeze_c, 1) # 该卷积的输出的squeeze_c是输入input_c的1/4self.fc2 nn.Conv2d(squeeze_c, input_c, 1)def forward(self, x: Tensor) - Tensor:scale F.adaptive_avg_pool2d(x, output_size(1, 1)) # 将特征矩阵每一个channel上的数据给平均池化到1x1的大小scale self.fc1(scale)scale F.relu(scale, inplaceTrue)scale self.fc2(scale)scale F.hardsigmoid(scale, inplaceTrue) # 激活函数return scale * x # 将得到的数据与传入的对应channel数据进行相乘InvertedResidual模块代码
# 定义block模块
# 此为block模块其包含第一个1x1卷积层、DeptWis卷积层、SE注意力机制层判断是否需求、第二个1x1卷积层、激活函数需要判断是否是非线性激活
class InvertedResidual(nn.Module):def __init__(self,cnf: InvertedResidualConfig, # cnf:配置类参数norm_layer: Callable[..., nn.Module]): # norm_layer# BN层super(InvertedResidual, self).__init__()if cnf.stride not in [1, 2]: # 判断某一层的配置文件其步长是否满足条件raise ValueError(illegal stride value.)# 判断是否进行短连接self.use_res_connect (cnf.stride 1 and cnf.input_c cnf.out_c) # 只有当步长为1并且输入通道等于输出通道数layers: List[nn.Module] []activation_layer nn.Hardswish if cnf.use_hs else nn.ReLU # 判断当前的激活函数类型# expand# 判断是否相等如果相等则不适用1x1的卷积层增加channel维度不相等的话才使用该层进行升维度if cnf.expanded_c ! cnf.input_c:layers.append(ConvBNActivation(cnf.input_c,cnf.expanded_c,kernel_size1,norm_layernorm_layer,activation_layeractivation_layer))# depthwiselayers.append(ConvBNActivation(cnf.expanded_c,cnf.expanded_c,kernel_sizecnf.kernel, # depthwise卷积的卷积核大小stridecnf.stride,groupscnf.expanded_c,norm_layernorm_layer, # BN层activation_layeractivation_layer))# 判断是否需要添加SE模块if cnf.use_se:layers.append(SqueezeExcitation(cnf.expanded_c))# projectlayers.append(ConvBNActivation(cnf.expanded_c,cnf.out_c,kernel_size1,norm_layernorm_layer, # BN 层activation_layernn.Identity)) # 此层的activation_layer就是进行里普通的线性激活没有做任何的处理self.block nn.Sequential(*layers)self.out_channels cnf.out_cself.is_strided cnf.stride 1def forward(self, x: Tensor) - Tensor:result self.block(x)if self.use_res_connect:result x # 进行shortcut连接return result整体代码
from typing import Callable, List, Optionalimport torch
from torch import nn, Tensor
from torch.nn import functional as F
from functools import partial# 得到同传入数据最近的8的整数倍数值
def _make_divisible(ch, divisor8, min_chNone):This function is taken from the original tf repo.It ensures that all layers have a channel number that is divisible by 8It can be seen here:https://github.com/tensorflow/models/blob/master/research/slim/nets/mobilenet/mobilenet.pyif min_ch is None:min_ch divisornew_ch max(min_ch, int(ch divisor / 2) // divisor * divisor)# Make sure that round down does not go down by more than 10%.if new_ch 0.9 * ch:new_ch divisorreturn new_ch# 普通卷积、BN、激活层模块
class ConvBNActivation(nn.Sequential):def __init__(self,in_planes: int, # 输入特征矩阵的通道out_planes: int, # 输出特征矩阵的通道kernel_size: int 3,stride: int 1,groups: int 1,norm_layer: Optional[Callable[..., nn.Module]] None, # 在卷积后的BN层activation_layer: Optional[Callable[..., nn.Module]] None): # 激活函数padding (kernel_size - 1) // 2if norm_layer is None:norm_layer nn.BatchNorm2dif activation_layer is None:activation_layer nn.ReLU6super(ConvBNActivation, self).__init__(nn.Conv2d(in_channelsin_planes,out_channelsout_planes,kernel_sizekernel_size,stridestride,paddingpadding,groupsgroups,biasFalse),norm_layer(out_planes), # BN层activation_layer(inplaceTrue))# 注意力机制模块SE模块即两个全连接层 该模块的基本流程是先进行自适应平均池化(1x1)———1x1的卷积层———relu激活层———1x1的卷积池化———hardsigmoid()激活函数激活
class SqueezeExcitation(nn.Module):def __init__(self, input_c: int, squeeze_factor: int 4):super(SqueezeExcitation, self).__init__()squeeze_c _make_divisible(input_c // squeeze_factor, 8) # 获得距离该数最近的8的整数倍的数字self.fc1 nn.Conv2d(input_c, squeeze_c, 1) # 该卷积的输出的squeeze_c是输入input_c的1/4 其作用与全连接层一样self.fc2 nn.Conv2d(squeeze_c, input_c, 1)def forward(self, x: Tensor) - Tensor:scale F.adaptive_avg_pool2d(x, output_size(1, 1)) # 将特征矩阵每一个channel上的数据给平均池化到1x1的大小scale self.fc1(scale)scale F.relu(scale, inplaceTrue)scale self.fc2(scale)scale F.hardsigmoid(scale, inplaceTrue) # 激活函数return scale * x # 将得到的数据与传入的对应channel数据进行相乘# 定义block的配置类
class InvertedResidualConfig:def __init__(self,input_c: int, # block模块中的第一个1x1卷积层的输入channel数kernel: int, # depthwise卷积的卷积核大小expanded_c: int, # block模块中的第一个1x1卷积层的输出channel数out_c: int, # 经过block模块中第二个1x1卷积层处理过后得到的channel数use_se: bool, # 是否使用注意力机制模块activation: str, # 激活方式stride: int, # 步长width_multi: float): # width_multi调节每个卷积层所使用channel的倍率因子self.input_c self.adjust_channels(input_c, width_multi)self.kernel kernelself.expanded_c self.adjust_channels(expanded_c, width_multi)self.out_c self.adjust_channels(out_c, width_multi)self.use_se use_seself.use_hs activation HS # whether using h-swish activationself.stride stridestaticmethoddef adjust_channels(channels: int, width_multi: float):return _make_divisible(channels * width_multi, 8)# 定义block模块
# 此为block模块其包含第一个1x1卷积层、DeptWis卷积层、SE注意力机制层判断是否需求、第二个1x1卷积层、激活函数需要判断是否是非线性激活
class InvertedResidual(nn.Module):def __init__(self,cnf: InvertedResidualConfig, # cnf:配置类参数norm_layer: Callable[..., nn.Module]): # norm_layer# BN层super(InvertedResidual, self).__init__()if cnf.stride not in [1, 2]: # 判断某一层的配置文件其步长是否满足条件raise ValueError(illegal stride value.)# 判断是否进行短连接self.use_res_connect (cnf.stride 1 and cnf.input_c cnf.out_c) # 只有当步长为1并且输入通道等于输出通道数layers: List[nn.Module] []activation_layer nn.Hardswish if cnf.use_hs else nn.ReLU # 判断当前的激活函数类型# expand# 判断是否相等如果相等则不适用1x1的卷积层增加channel维度不相等的话才使用该层进行升维度if cnf.expanded_c ! cnf.input_c:layers.append(ConvBNActivation(cnf.input_c,cnf.expanded_c,kernel_size1,norm_layernorm_layer,activation_layeractivation_layer))# depthwiselayers.append(ConvBNActivation(cnf.expanded_c,cnf.expanded_c,kernel_sizecnf.kernel, # depthwise卷积的卷积核大小stridecnf.stride,groupscnf.expanded_c, # 深度DW卷积norm_layernorm_layer, # BN层activation_layeractivation_layer))# 判断是否需要添加SE模块if cnf.use_se:layers.append(SqueezeExcitation(cnf.expanded_c))# projectlayers.append(ConvBNActivation(cnf.expanded_c,cnf.out_c,kernel_size1,norm_layernorm_layer, # BN 层activation_layernn.Identity)) # 此层的activation_layer就是进行里普通的线性激活没有做任何的处理self.block nn.Sequential(*layers)self.out_channels cnf.out_cself.is_strided cnf.stride 1def forward(self, x: Tensor) - Tensor:result self.block(x)if self.use_res_connect:result x # 进行shortcut连接return result# MobileNetV3网络结构基础框架其包括模型的第一层卷积层———nx【bneckBlock模块】———1x1的卷积层———自适应平均池化层———全连接层———全连接层
class MobileNetV3(nn.Module):def __init__(self,inverted_residual_setting: List[InvertedResidualConfig], # beneckBlock结构一系列参数列表last_channel: int, # 对应的是倒数第二个全连接层输出节点数 1280num_classes: int 1000, # 类别个数block: Optional[Callable[..., nn.Module]] None, # InvertedResidual核心模块norm_layer: Optional[Callable[..., nn.Module]] None):super(MobileNetV3, self).__init__()if not inverted_residual_setting:raise ValueError(The inverted_residual_setting should not be empty.)elif not (isinstance(inverted_residual_setting, List) andall([isinstance(s, InvertedResidualConfig) for s in inverted_residual_setting])):raise TypeError(The inverted_residual_setting should be List[InvertedResidualConfig])if block is None:block InvertedResidual # block类if norm_layer is None:norm_layer partial(nn.BatchNorm2d, eps0.001, momentum0.01) # partial()为python方法即为nn.BatchNorm2d传入默认的两个参数layers: List[nn.Module] []# building first layer# 构建第一层卷积结构firstconv_output_c inverted_residual_setting[0].input_c # 表示第一个卷积层输出的channel数layers.append(ConvBNActivation(3, # 输入图像数据的channel数firstconv_output_c, # 输出channelkernel_size3,stride2,norm_layernorm_layer,activation_layernn.Hardswish))# building inverted residual blocks# 利用循环的方式添加block模块将每层的配置文件传给blockfor cnf in inverted_residual_setting:layers.append(block(cnf, norm_layer))# building last several layerslastconv_input_c inverted_residual_setting[-1].out_c # 最后的bneckblock的输出channellastconv_output_c 6 * lastconv_input_c # lastconv_output_c 与 最后的bneckblock的输出channel数是六倍的关系# 定义最后一层的卷积层layers.append(ConvBNActivation(lastconv_input_c, # 最后的bneckblock的输出channel数lastconv_output_c, # lastconv_output_c 与 最后的bneckblock的输出channel数是六倍的关系kernel_size1,norm_layernorm_layer,activation_layernn.Hardswish))self.features nn.Sequential(*layers)self.avgpool nn.AdaptiveAvgPool2d(1)self.classifier nn.Sequential(nn.Linear(lastconv_output_c, last_channel),nn.Hardswish(inplaceTrue),nn.Dropout(p0.2, inplaceTrue),nn.Linear(last_channel, num_classes))# initial weightsfor m in self.modules():if isinstance(m, nn.Conv2d):nn.init.kaiming_normal_(m.weight, modefan_out)if m.bias is not None:nn.init.zeros_(m.bias)elif isinstance(m, (nn.BatchNorm2d, nn.GroupNorm)):nn.init.ones_(m.weight)nn.init.zeros_(m.bias)elif isinstance(m, nn.Linear):nn.init.normal_(m.weight, 0, 0.01)nn.init.zeros_(m.bias)def _forward_impl(self, x: Tensor) - Tensor:x self.features(x)x self.avgpool(x)x torch.flatten(x, 1)x self.classifier(x)return xdef forward(self, x: Tensor) - Tensor:return self._forward_impl(x)### 构建large基础mobilenet_v3_large模型
def mobilenet_v3_large(num_classes: int 1000,reduced_tail: bool False) - MobileNetV3:Constructs a large MobileNetV3 architecture fromSearching for MobileNetV3 https://arxiv.org/abs/1905.02244.weights_link:https://download.pytorch.org/models/mobilenet_v3_large-8738ca79.pthArgs:num_classes (int): number of classesreduced_tail (bool): If True, reduces the channel counts of all feature layersbetween C4 and C5 by 2. It is used to reduce the channel redundancy in thebackbone for Detection and Segmentation.width_multi 1.0bneck_conf partial(InvertedResidualConfig, width_multiwidth_multi)adjust_channels partial(InvertedResidualConfig.adjust_channels, width_multiwidth_multi)reduce_divider 2 if reduced_tail else 1 # 是否较少网络参数标志默认是False即不减少# # beneckBlock结构一系列参数列表inverted_residual_setting [# input_c, kernel, expanded_c, out_c, use_se, activation, stridebneck_conf(16, 3, 16, 16, False, RE, 1),bneck_conf(16, 3, 64, 24, False, RE, 2), # C1bneck_conf(24, 3, 72, 24, False, RE, 1),bneck_conf(24, 5, 72, 40, True, RE, 2), # C2bneck_conf(40, 5, 120, 40, True, RE, 1),bneck_conf(40, 5, 120, 40, True, RE, 1),bneck_conf(40, 3, 240, 80, False, HS, 2), # C3bneck_conf(80, 3, 200, 80, False, HS, 1),bneck_conf(80, 3, 184, 80, False, HS, 1),bneck_conf(80, 3, 184, 80, False, HS, 1),bneck_conf(80, 3, 480, 112, True, HS, 1),bneck_conf(112, 3, 672, 112, True, HS, 1),bneck_conf(112, 5, 672, 160 // reduce_divider, True, HS, 2), # C4bneck_conf(160 // reduce_divider, 5, 960 // reduce_divider, 160 // reduce_divider, True, HS, 1),bneck_conf(160 // reduce_divider, 5, 960 // reduce_divider, 160 // reduce_divider, True, HS, 1),]last_channel adjust_channels(1280 // reduce_divider) # C5return MobileNetV3(inverted_residual_settinginverted_residual_setting,last_channellast_channel,num_classesnum_classes)### 构建small基础mobilenet_v3_small模型
def mobilenet_v3_small(num_classes: int 1000,reduced_tail: bool False) - MobileNetV3:Constructs a large MobileNetV3 architecture fromSearching for MobileNetV3 https://arxiv.org/abs/1905.02244.weights_link:https://download.pytorch.org/models/mobilenet_v3_small-047dcff4.pthArgs:num_classes (int): number of classesreduced_tail (bool): If True, reduces the channel counts of all feature layersbetween C4 and C5 by 2. It is used to reduce the channel redundancy in thebackbone for Detection and Segmentation.width_multi 1.0bneck_conf partial(InvertedResidualConfig, width_multiwidth_multi)adjust_channels partial(InvertedResidualConfig.adjust_channels, width_multiwidth_multi)reduce_divider 2 if reduced_tail else 1inverted_residual_setting [# input_c, kernel, expanded_c, out_c, use_se, activation, stridebneck_conf(16, 3, 16, 16, True, RE, 2), # C1bneck_conf(16, 3, 72, 24, False, RE, 2), # C2bneck_conf(24, 3, 88, 24, False, RE, 1),bneck_conf(24, 5, 96, 40, True, HS, 2), # C3bneck_conf(40, 5, 240, 40, True, HS, 1),bneck_conf(40, 5, 240, 40, True, HS, 1),bneck_conf(40, 5, 120, 48, True, HS, 1),bneck_conf(48, 5, 144, 48, True, HS, 1),bneck_conf(48, 5, 288, 96 // reduce_divider, True, HS, 2), # C4bneck_conf(96 // reduce_divider, 5, 576 // reduce_divider, 96 // reduce_divider, True, HS, 1),bneck_conf(96 // reduce_divider, 5, 576 // reduce_divider, 96 // reduce_divider, True, HS, 1)]last_channel adjust_channels(1024 // reduce_divider) # C5return MobileNetV3(inverted_residual_settinginverted_residual_setting,last_channellast_channel,num_classesnum_classes)pytorch代码复现MobileNet V1~V2
本项目包含训练MobileNet V1、V2、V2模型
项目目录 项目代码下载地址 项目代码下载地址
