ResNet学习笔记

https://www.bilibili.com/video/BV1T7411T7wa
分析ResNet网络的结构，了解Batch Normalization和迁移学习的相关概念

不能简单堆叠cnn与pool：效果差了
1.梯度消失与爆炸明显
消失：

解决：数据标准化处理，模型初始化，以及BatchNormalization
2.退化的问题
用残差结构解决

两种残差结构：左面的给小的，右边的给大的

7：35 很详细

下采样（用右边的卷积核降维）

迁移学习：
31：58

model

import torch.nn as nn
import torchclass BasicBlock(nn.Module):expansion = 1#stride，实线为1 虚线为2 05：04def __init__(self, in_channel, out_channel, stride=1, downsample=None):super(BasicBlock, self).__init__()self.conv1 = nn.Conv2d(in_channels=in_channel, out_channels=out_channel,kernel_size=3, stride=stride, padding=1, bias=False)#bias 不需要使用偏置self.bn1 = nn.BatchNorm2d(out_channel)self.relu = nn.ReLU()self.conv2 = nn.Conv2d(in_channels=out_channel, out_channels=out_channel,kernel_size=3, stride=1, padding=1, bias=False)self.bn2 = nn.BatchNorm2d(out_channel)self.downsample = downsample#对应虚线的残差结构def forward(self, x):identity = x#捷径的输出值if self.downsample is not None:identity = self.downsample(x)out = self.conv1(x)out = self.bn1(out)out = self.relu(out)out = self.conv2(out)out = self.bn2(out)out += identityout = self.relu(out)return outclass Bottleneck(nn.Module):#50，101，152的残差结构expansion = 4#4倍def __init__(self, in_channel, out_channel, stride=1, downsample=None):super(Bottleneck, self).__init__()self.conv1 = nn.Conv2d(in_channels=in_channel, out_channels=out_channel,kernel_size=1, stride=1, bias=False)  # squeeze channelsself.bn1 = nn.BatchNorm2d(out_channel)# -----------------------------------------self.conv2 = nn.Conv2d(in_channels=out_channel, out_channels=out_channel,kernel_size=3, stride=stride, bias=False, padding=1)self.bn2 = nn.BatchNorm2d(out_channel)# -----------------------------------------self.conv3 = nn.Conv2d(in_channels=out_channel, out_channels=out_channel*self.expansion,kernel_size=1, stride=1, bias=False)  # unsqueeze channels#out*expansion = 4#4倍self.bn3 = nn.BatchNorm2d(out_channel*self.expansion)self.relu = nn.ReLU(inplace=True)self.downsample = downsampledef forward(self, x):identity = xif self.downsample is not None:identity = self.downsample(x)out = self.conv1(x)out = self.bn1(out)out = self.relu(out)out = self.conv2(out)out = self.bn2(out)out = self.relu(out)out = self.conv3(out)out = self.bn3(out)out += identityout = self.relu(out)return outclass ResNet(nn.Module):def __init__(self, block, blocks_num, num_classes=1000, include_top=True):super(ResNet, self).__init__()self.include_top = include_topself.in_channel = 64self.conv1 = nn.Conv2d(3, self.in_channel, kernel_size=7, stride=2,padding=3, bias=False)self.bn1 = nn.BatchNorm2d(self.in_channel)self.relu = nn.ReLU(inplace=True)self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1)self.layer1 = self._make_layer(block, 64, blocks_num[0])self.layer2 = self._make_layer(block, 128, blocks_num[1], stride=2)self.layer3 = self._make_layer(block, 256, blocks_num[2], stride=2)self.layer4 = self._make_layer(block, 512, blocks_num[3], stride=2)if self.include_top:self.avgpool = nn.AdaptiveAvgPool2d((1, 1))  # output size = (1, 1)self.fc = nn.Linear(512 * block.expansion, num_classes)for m in self.modules():if isinstance(m, nn.Conv2d):nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu')def _make_layer(self, block, channel, block_num, stride=1):downsample = Noneif stride != 1 or self.in_channel != channel * block.expansion:downsample = nn.Sequential(nn.Conv2d(self.in_channel, channel * block.expansion, kernel_size=1, stride=stride, bias=False),nn.BatchNorm2d(channel * block.expansion))layers = []layers.append(block(self.in_channel, channel, downsample=downsample, stride=stride))self.in_channel = channel * block.expansionfor _ in range(1, block_num):layers.append(block(self.in_channel, channel))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)if self.include_top:x = self.avgpool(x)x = torch.flatten(x, 1)x = self.fc(x)return xdef resnet34(num_classes=2, include_top=True):return ResNet(BasicBlock, [3, 4, 6, 3], num_classes=num_classes, include_top=include_top)def resnet101(num_classes=2, include_top=True):return ResNet(Bottleneck, [3, 4, 23, 3], num_classes=num_classes, include_top=include_top)

train


net = resnet34()
# load pretrain weights
# model_weight_path = "./resnet34-pre.pth"
# missing_keys, unexpected_keys = net.load_state_dict(torch.load(model_weight_path), strict=False)
# for param in net.parameters():
#     param.requires_grad = False
# change fc layer structure
inchannel = net.fc.in_features
net.fc = nn.Linear(inchannel, 2)
net.to(device)loss_function = nn.CrossEntropyLoss()
optimizer = optim.Adam(net.parameters(), lr=0.0001)

ResNet学习笔记相关推荐

ResNet网络学习笔记。
ResNet网络学习看b站霹雳吧啦Wz 的视频总结的学习笔记! 视频的地址大佬的Github代码 1.ResNet详解 ResNet 网络是在2015年由微软实验室提出,斩获当年 ImageNe ...
resnet50网络结构_学习笔记（一）：分析resnet源码理解resnet网络结构
最近在跑实验的过程中一直在使用resnet50和resnet34,为了弄清楚网络的结构和原理的实现,打开resnet的源码进行了学习. 残差网络学习的原理针对神经网络过深而导致的学习准确率饱和甚至是 ...
ResNet 小白学习笔记
写在前面直接看论文对我这个小白来说太不友好了,几次放弃 /(ㄒoㄒ)/~~ .幸好找到了一个通俗易懂的视频:6.1 ResNet网络结构,BN以及迁移学习详解,以下笔记大部分基于视频内容,再补充了一 ...
深度学习入门之PyTorch学习笔记：卷积神经网络
深度学习入门之PyTorch学习笔记绪论 1 深度学习介绍 2 深度学习框架 3 多层全连接网络 4 卷积神经网络 4.1 主要任务及起源 4.2 卷积神经网络的原理和结构 4.2.1 卷积层 1. ...
AI学习笔记（十四）CNN之图像分割
AI学习笔记之CNN之图像分割图像分割问题引入实现技术手段及分类语义分割-FCN(Fully Convolutional Networks) FCN--deconv 反卷积的具体步骤 Unpo ...
AI学习笔记（十一）CNN之图像识别(上）
AI学习笔记之CNN之图像识别(上) 图像识别图像识别简介模式识别图像识别的过程图像识别的应用分类与检测 VGG Resnet 迁移学习&inception 卷积神经网络迁移学习fi ...
AI学习笔记（十一）CNN之图像识别(下）
AI学习笔记之CNN之图像识别(下) 图像识别图像识别简介模式识别图像识别的过程图像识别的应用分类与检测 VGG Resnet 迁移学习&inception 卷积神经网络迁移学习fi ...
AI学习笔记（十二）物体检测（上）
AI学习笔记之物体检测(上) 物体检测简介常见检测网络 IOU TP.TN.FP.FN precision(精确度)和recall(召回率) 边框回归边框回归具体方法 Selective Sear ...
AI学习笔记（十）卷积神经网络
AI学习笔记之卷积神经网络卷积神经网络简介卷积层池化层卷积神经网络卷积核填充padding 常见的卷积神经网络 cifar-10预测实例数据预处理--图像增强图像增强常用方法 Alex ...

ResNet学习笔记

ResNet学习笔记相关推荐

最新文章

热门文章