resnet50 cifar

ResNet50在CIFAR数据集上的应用

ResNet架构通过引入残差学习单元解决了深层神经网络中的退化问题，使得训练更深的网络成为可能[^2]。对于像CIFAR这样的小型图像数据集，尽管原始论文主要关注ImageNet数据集，但ResNet同样可以应用于较小规模的任务。

以下是基于PyTorch框架的一个简化版本ResNet50模型，在CIFAR数据集上的实现：

import torch.nn as nn
import torch.nn.functional as F


class BasicBlock(nn.Module):
    expansion = 1
    
    def __init__(self, in_planes, planes, stride=1):
        super(BasicBlock, self).__init__()
        self.conv1 = nn.Conv2d(in_planes, planes, kernel_size=3, stride=stride, padding=1, bias=False)
        self.bn1 = nn.BatchNorm2d(planes)
        self.conv2 = nn.Conv2d(planes, planes, kernel_size=3, stride=1, padding=1, bias=False)
        self.bn2 = nn.BatchNorm2d(planes)

        self.shortcut = nn.Sequential()
        if stride != 1 or in_planes != self.expansion*planes:
            self.shortcut = nn.Sequential(
                nn.Conv2d(in_planes, self.expansion*planes, kernel_size=1, stride=stride, bias=False),
                nn.BatchNorm2d(self.expansion*planes)
            )

    def forward(self, x):
        out = F.relu(self.bn1(self.conv1(x)))
        out = self.bn2(self.conv2(out))
        out += self.shortcut(x)
        out = F.relu(out)
        return out


def make_layer(block, in_channels, channels, num_blocks, stride):
    strides = [stride] + [1]*(num_blocks-1)
    layers = []
    for stride in strides:
        layers.append(block(in_channels, channels, stride))
        in_channels = block.expansion * channels
    return nn.Sequential(*layers)


class ResNet_CIFAR(nn.Module):
    def __init__(self, block, num_blocks, num_classes=10):
        super(ResNet_CIFAR, self).__init__()
        self.in_planes = 64
        
        self.conv1 = nn.Conv2d(3, 64, kernel_size=3, stride=1, padding=1, bias=False)
        self.bn1 = nn.BatchNorm2d(64)
        
        # Adjusted to fit the smaller input size of CIFAR images (32x32 vs ImageNet's 224x224).
        self.layer1 = make_layer(block, 64, 64, num_blocks[0], stride=1)
        self.layer2 = make_layer(block, 64, 128, num_blocks[1], stride=2)
        self.layer3 = make_layer(block, 128, 256, num_blocks[2], stride=2)
        self.layer4 = make_layer(block, 256, 512, num_blocks[3], stride=2)
        
        self.linear = nn.Linear(512*block.expansion, num_classes)

    def forward(self, x):
        out = F.relu(self.bn1(self.conv1(x)))
        out = self.layer1(out)
        out = self.layer2(out)
        out = self.layer3(out)
        out = self.layer4(out)
        out = F.avg_pool2d(out, 4)
        out = out.view(out.size(0), -1)
        out = self.linear(out)
        return out


def resnet_cifar():
    """Constructs a modified ResNet-50 model suitable for CIFAR."""
    return ResNet_CIFAR(Bottleneck, [3, 4, 6, 3])