#!/usr/bin/env python # encoding: utf-8 ''' @author: MingDong @file: resnet.py @desc: Original ResNet model, including ResNet18, ResNet34, ResNet50, ResNet101 and ResNet152, we removed the last global average pooling layer and replaced it with a fully connected layer with dimension of 512. BN is used for fast convergence. ''' import torch import torch.nn as nn def ResNet18(): model = ResNet(BasicBlock, [2, 2, 2, 2]) return model def ResNet34(): model = ResNet(BasicBlock, [3, 4, 6, 3]) return model def ResNet50(): model = ResNet(Bottleneck, [3, 4, 6, 3]) return model def ResNet101(): model = ResNet(Bottleneck, [3, 4, 23, 3]) return model def ResNet152(): model = ResNet(Bottleneck, [3, 8, 36, 3]) return model __all__ = ['ResNet', 'ResNet18', 'ResNet34', 'ResNet50', 'ResNet101', 'ResNet152'] 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) 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): super(BasicBlock, self).__init__() self.conv1 = nn.Conv2d(inplanes, planes, kernel_size=3, stride=stride, padding=1, bias=False) self.bn1 = nn.BatchNorm2d(planes) self.relu1 = nn.ReLU(inplace=True) self.conv2 = nn.Conv2d(planes, planes, kernel_size=3, stride=1, padding=1, bias=False) self.bn2 = nn.BatchNorm2d(planes) self.downsample = downsample self.stride = stride self.skip_add = nn.quantized.FloatFunctional() # Remember to use two independent ReLU for layer fusion. self.relu2 = nn.ReLU(inplace=True) def forward(self, x): identity = x out = self.conv1(x) out = self.bn1(out) out = self.relu1(out) out = self.conv2(out) out = self.bn2(out) if self.downsample is not None: identity = self.downsample(x) # Use FloatFunctional for addition for quantization compatibility # out += identity # out = torch.add(identity, out) out = self.skip_add.add(identity, out) out = self.relu2(out) return out class Bottleneck(nn.Module): expansion = 4 def __init__(self, inplanes, planes, stride=1, downsample=None): super(Bottleneck, self).__init__() self.conv1 = nn.Conv2d(inplanes, planes, kernel_size=1, stride=1, bias=False) self.bn1 = nn.BatchNorm2d(planes) self.relu1 = nn.ReLU(inplace=True) self.conv2 = nn.Conv2d(planes, planes, kernel_size=3, stride=stride, padding=1, bias=False) self.bn2 = nn.BatchNorm2d(planes) self.relu2 = nn.ReLU(inplace=True) self.conv3 = nn.Conv2d(planes, planes * self.expansion, kernel_size=1, stride=1, bias=False) self.bn3 = nn.BatchNorm2d(planes * self.expansion) self.downsample = downsample self.stride = stride self.skip_add = nn.quantized.FloatFunctional() self.relu3 = nn.ReLU(inplace=True) def forward(self, x): identity = x out = self.conv1(x) out = self.bn1(out) out = self.relu1(out) out = self.conv2(out) out = self.bn2(out) out = self.relu2(out) out = self.conv3(out) out = self.bn3(out) if self.downsample is not None: identity = self.downsample(x) # out += identity # out = torch.add(identity, out) out = self.skip_add.add(identity, out) out = self.relu3(out) return out class Flatten(nn.Module): def forward(self, x): # return input.view(input.size(0), -1) x = x.reshape(x.size(0), -1) return torch.unsqueeze(torch.unsqueeze(x, 2), 3) class ResNet(nn.Module): def __init__(self, block, layers, feature_dim=512, drop_ratio=0.4, zero_init_residual=False): super(ResNet, self).__init__() self.inplanes = 64 self.conv1 = nn.Conv2d(3, 64, kernel_size=3, stride=1, padding=1, bias=False) self.bn1 = nn.BatchNorm2d(64) 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) self.layer3 = self._make_layer(block, 256, layers[2], stride=2) self.layer4 = self._make_layer(block, 512, layers[3], stride=2) self.output_layer = nn.Sequential(nn.BatchNorm2d(512 * block.expansion), nn.Dropout(drop_ratio), Flatten(), nn.Conv2d(512 * block.expansion * 7 * 7, feature_dim, 1), nn.BatchNorm2d(feature_dim), nn.Flatten()) # self.output_bn2d = nn.BatchNorm2d(512 * block.expansion) # self.output_drop = nn.Dropout(drop_ratio) # self.output_linear = nn.Linear(512 * block.expansion * 7 * 7, feature_dim) # self.output_bn1d = nn.BatchNorm1d(feature_dim) # # self.avgpool = nn.AdaptiveAvgPool2d((1, 1)) # self.fc = nn.Linear(512 * block.expansion, feature_dim) 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.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 checkpoints 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): downsample = None if stride != 1 or self.inplanes != planes * block.expansion: downsample = nn.Sequential( conv1x1(self.inplanes, planes * block.expansion, stride), nn.BatchNorm2d(planes * block.expansion), ) layers = [] layers.append(block(self.inplanes, planes, stride, downsample)) self.inplanes = planes * block.expansion for _ in range(1, blocks): layers.append(block(self.inplanes, planes)) 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.output_layer(x) # x = self.output_bn2d(x) # x = self.output_drop(x) # x = torch.flatten(x, 1) # x = self.output_linear(x) # x = self.output_bn1d(x) return x if __name__ == "__main__": x = torch.Tensor(2, 3, 112, 112) net = ResNet50() print(net) x = net(x) print(x.shape)