mobilenetv3代码

以下是使用PyTorch实现MobileNetV3的代码： ```python import torch import torch.nn as nn import torch.nn.functional as F class Hswish(nn.Module): def __init__(self, inplace=True): super(Hswish, self).__init__() self.inplace = inplace def forward(self, x): if self.inplace: return x.mul_(F.relu6(x + 3., inplace=True)) / 6. else: return F.relu6(x + 3.) * x / 6. class Hsigmoid(nn.Module): def __init__(self, inplace=True): super(Hsigmoid, self).__init__() self.inplace = inplace def forward(self, x): if self.inplace: return F.relu6(x + 3., inplace=True) / 6. else: return F.relu6(x + 3.) / 6. class SEModule(nn.Module): def __init__(self, in_channels, reduction_ratio=4): super(SEModule, self).__init__() self.avg_pool = nn.AdaptiveAvgPool2d(1) self.fc1 = nn.Conv2d(in_channels, in_channels // reduction_ratio, kernel_size=1, bias=False) self.relu = nn.ReLU(inplace=True) self.fc2 = nn.Conv2d(in_channels // reduction_ratio, in_channels, kernel_size=1, bias=False) self.hsigmoid = Hsigmoid() def forward(self, x): module_input = x x = self.avg_pool(x) x = self.fc1(x) x = self.relu(x) x = self.fc2(x) x = self.hsigmoid(x) return module_input * x class MobileNetV3Block(nn.Module): def __init__(self, in_channels, out_channels, kernel_size, stride, use_se, activation): super(MobileNetV3Block, self).__init__() self.use_se = use_se self.activation = activation padding = (kernel_size - 1) // 2 self.conv1 = nn.Conv2d(in_channels, in_channels, kernel_size=1, stride=1, padding=0, bias=False) self.bn1 = nn.BatchNorm2d(in_channels) self.conv2 = nn.Conv2d(in_channels, in_channels, kernel_size=kernel_size, stride=stride, padding=padding, groups=in_channels, bias=False) self.bn2 = nn.BatchNorm2d(in_channels) self.conv3 = nn.Conv2d(in_channels, out_channels, kernel_size=1, stride=1, padding=0, bias=False) self.bn3 = nn.BatchNorm2d(out_channels) if use_se: self.se = SEModule(out_channels) if activation == 'relu': self.activation_fn = nn.ReLU(inplace=True) elif activation == 'hswish': self.activation_fn = Hswish(inplace=True) def forward(self, x): module_input = x x = self.conv1(x) x = self.bn1(x) x = self.activation_fn(x) x = self.conv2(x) x = self.bn2(x) x = self.activation_fn(x) x = self.conv3(x) x = self.bn3(x) if self.use_se: x = self.se(x) x += module_input return x class MobileNetV3Large(nn.Module): def __init__(self, num_classes=1000): super(MobileNetV3Large, self).__init__() # Settings for feature extraction part self.conv1 = nn.Conv2d(3, 16, kernel_size=3, stride=2, padding=1, bias=False) self.bn1 = nn.BatchNorm2d(16) self.hs1 = Hswish() self.block1 = MobileNetV3Block(16, 16, kernel_size=3, stride=1, use_se=False, activation='relu') self.block2 = MobileNetV3Block(16, 24, kernel_size=3, stride=2, use_se=False, activation='relu') self.block3 = MobileNetV3Block(24, 24, kernel_size=3, stride=1, use_se=False, activation='relu') self.block4 = MobileNetV3Block(24, 40, kernel_size=5, stride=2, use_se=True, activation='relu') self.block5 = MobileNetV3Block(40, 40, kernel_size=5, stride=1, use_se=True, activation='relu') self.block6 = MobileNetV3Block(40, 40, kernel_size=5, stride=1, use_se=True, activation='relu') self.block7 = MobileNetV3Block(40, 80, kernel_size=3, stride=2, use_se=False, activation='hswish') self.block8 = MobileNetV3Block(80, 80, kernel_size=3, stride=1, use_se=False, activation='hswish') self.block9 = MobileNetV3Block(80, 80, kernel_size=3, stride=1, use_se=False, activation='hswish') self.block10 = MobileNetV3Block(80, 112, kernel_size=3, stride=1, use_se=True, activation='hswish') self.block11 = MobileNetV3Block(112, 112, kernel_size=3, stride=1, use_se=True, activation='hswish') self.block12 = MobileNetV3Block(112, 160, kernel_size=5, stride=2, use_se=True, activation='hswish') self.block13 = MobileNetV3Block(160, 160, kernel_size=5, stride=1, use_se=True, activation='hswish') self.block14 = MobileNetV3Block(160, 160, kernel_size=5, stride=1, use_se=True, activation='hswish') # Settings for classification part self.conv2 = nn.Conv2d(160, 960, kernel_size=1, stride=1, padding=0, bias=False) self.bn2 = nn.BatchNorm2d(960) self.hs2 = Hswish() self.avgpool = nn.AdaptiveAvgPool2d(1) self.dropout = nn.Dropout(p=0.2, inplace=True) self.fc = nn.Linear(960, num_classes) # Weight initialization self._initialize_weights() def forward(self, x): x = self.conv1(x) x = self.bn1(x) x = self.hs1(x) x = self.block1(x) x = self.block2(x) x = self.block3(x) x = self.block4(x) x = self.block5(x) x = self.block6(x) x = self.block7(x) x = self.block8(x) x = self.block9(x) x = self.block10(x) x = self.block11(x) x = self.block12(x) x = self.block13(x) x = self.block14(x) x = self.conv2(x) x = self.bn2(x) x = self.hs2(x) x = self.avgpool(x) x = x.view(x.size(0), -1) x = self.dropout(x) x = self.fc(x) return x def _initialize_weights(self): for m in self.modules(): if isinstance(m, nn.Conv2d): nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu') 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) nn.init.zeros_(m.bias) ``` 这是MobileNetV3的大型版本，你也可以根据需要修改输入和输出通道数等参数来创建其他版本的MobileNetV3。

### MobilenetV3 的代码实现 #### 使用 PyTorch 实现的 MobileNetV3 结构如下： ```python import torch.nn as nn import torch class h_sigmoid(nn.Module): def __init__(self, inplace=True): super(h_sigmoid, self).__init__() self.relu = nn.ReLU6(inplace=inplace) def forward(self, x): return self.relu(x + 3) / 6 class h_swish(nn.Module): def __init__(self, inplace=True): super(h_swish, self).__init__() self.sigmoid = h_sigmoid(inplace=inplace) def forward(self, x): return x * self.sigmoid(x) class SELayer(nn.Module): def __init__(self, channel, reduction=4): super(SELayer, self).__init__() self.avg_pool = nn.AdaptiveAvgPool2d(1) self.fc = nn.Sequential( nn.Linear(channel, channel // reduction), nn.ReLU(inplace=True), nn.Linear(channel // reduction, channel), h_sigmoid() ) def forward(self, x): b, c, _, _ = x.size() y = self.avg_pool(x).view(b, c) y = self.fc(y).view(b, c, 1, 1) return x * y def conv_3x3_bn(inp, oup, stride): return nn.Sequential( nn.Conv2d(inp, oup, 3, stride, 1, bias=False), nn.BatchNorm2d(oup), h_swish() ) def conv_1x1_bn(inp, oup): return nn.Sequential( nn.Conv2d(inp, oup, 1, 1, 0, bias=False), nn.BatchNorm2d(oup), h_swish() ) class InvertedResidual(nn.Module): def __init__(self, inp, hidden_dim, oup, kernel_size, stride, use_se, use_hs): super(InvertedResidual, self).__init__() assert stride in [1, 2] self.identity = stride == 1 and inp == oup if inp == hidden_dim: self.conv = nn.Sequential( # dw nn.Conv2d(hidden_dim, hidden_dim, kernel_size, stride, (kernel_size - 1) // 2, groups=hidden_dim, bias=False), nn.BatchNorm2d(hidden_dim), h_swish() if use_hs else nn.ReLU(inplace=True), # Squeeze-and-Excite SELayer(hidden_dim) if use_se else nn.Identity(), # pw-linear nn.Conv2d(hidden_dim, oup, 1, 1, 0, bias=False), nn.BatchNorm2d(oup), ) else: self.conv = nn.Sequential( # pw nn.Conv2d(inp, hidden_dim, 1, 1, 0, bias=False), nn.BatchNorm2d(hidden_dim), h_swish() if use_hs else nn.ReLU(inplace=True), # dw nn.Conv2d(hidden_dim, hidden_dim, kernel_size, stride, (kernel_size - 1) // 2, groups=hidden_dim, bias=False), nn.BatchNorm2d(hidden_dim), # Squeeze-and-Excite SELayer(hidden_dim) if use_se else nn.Identity(), h_swish() if use_hs else nn.ReLU(inplace=True), # pw-linear nn.Conv2d(hidden_dim, oup, 1, 1, 0, bias=False), nn.BatchNorm2d(oup), ) def forward(self, x): if self.identity: return x + self.conv(x) else: return self.conv(x) class MobileNetV3(nn.Module): def __init__(cfgs, mode='small', num_classes=1000, width_mult=1.): super(MobileNetV3, self).__init__() # setting of inverted residual blocks self.cfgs = cfgs assert mode in ['large', 'small'] input_channel = _make_divisible(16 * width_mult, 8) layers = [conv_3x3_bn(3, input_channel, 2)] block = InvertedResidual for k, t, c, use_se, use_hs, s in self.cfgs: output_channel = _make_divisible(c * width_mult, 8) exp_size = _make_divisible(input_channel * t, 8) layers.append(block(input_channel, exp_size, output_channel, k, s, use_se, use_hs)) input_channel = output_channel self.features = nn.Sequential(*layers) self.classifier = nn.Sequential( nn.Linear(exp_size, exp_size), h_swish(), nn.Linear(exp_size, num_classes), ) def forward(self, x): x = self.features(x) x = x.mean([2, 3]) x = self.classifier(x) return x ``` 此代码展示了如何构建一个完整的 MobileNetV3 模型结构，包括自定义激活函数 `h-swish` 和注意力机制模块 SE-Layer。通过调整配置文件中的超参数设置，可以轻松切换到不同的网络变体[^1]。