diff --git a/.gitignore b/.gitignore new file mode 100644 index 0000000..a109725 --- /dev/null +++ b/.gitignore @@ -0,0 +1,164 @@ +# Byte-compiled / optimized / DLL files +__pycache__/ +*.py[cod] +*$py.class + +# C extensions +*.so + +# Distribution / packaging +CIFAR10 +model +logs +.idea +.Python +build/ +develop-eggs/ +dist/ +downloads/ +eggs/ +.eggs/ +lib/ +lib64/ +parts/ +sdist/ +var/ +wheels/ +share/python-wheels/ +*.egg-info/ +.installed.cfg +*.egg +MANIFEST + +# PyInstaller +# Usually these files are written by a python script from a template +# before PyInstaller builds the exe, so as to inject date/other infos into it. +*.manifest +*.spec + +# Installer logs +pip-log.txt +pip-delete-this-directory.txt + +# Unit test / coverage reports +htmlcov/ +.tox/ +.nox/ +.coverage +.coverage.* +.cache +nosetests.xml +coverage.xml +*.cover +*.py,cover +.hypothesis/ +.pytest_cache/ +cover/ + +# Translations +*.mo +*.pot + +# Django stuff: +*.log +local_settings.py +db.sqlite3 +db.sqlite3-journal + +# Flask stuff: +instance/ +.webassets-cache + +# Scrapy stuff: +.scrapy + +# Sphinx documentation +docs/_build/ + +# PyBuilder +.pybuilder/ +target/ + +# Jupyter Notebook +.ipynb_checkpoints + +# IPython +profile_default/ +ipython_config.py + +# pyenv +# For a library or package, you might want to ignore these files since the code is +# intended to run in multiple environments; otherwise, check them in: +# .python-version + +# pipenv +# According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control. +# However, in case of collaboration, if having platform-specific dependencies or dependencies +# having no cross-platform support, pipenv may install dependencies that don't work, or not +# install all needed dependencies. +#Pipfile.lock + +# poetry +# Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control. +# This is especially recommended for binary packages to ensure reproducibility, and is more +# commonly ignored for libraries. +# https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control +#poetry.lock + +# pdm +# Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control. +#pdm.lock +# pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it +# in version control. +# https://pdm.fming.dev/#use-with-ide +.pdm.toml + +# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm +__pypackages__/ + +# Celery stuff +celerybeat-schedule +celerybeat.pid + +# SageMath parsed files +*.sage.py + +# Environments +.env +.venv +env/ +venv/ +ENV/ +env.bak/ +venv.bak/ + +# Spyder project settings +.spyderproject +.spyproject + +# Rope project settings +.ropeproject + +# mkdocs documentation +/site + +# mypy +.mypy_cache/ +.dmypy.json +dmypy.json + +# Pyre type checker +.pyre/ + +# pytype static type analyzer +.pytype/ + +# Cython debug symbols +cython_debug/ + +# PyCharm +# JetBrains specific template is maintained in a separate JetBrains.gitignore that can +# be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore +# and can be added to the global gitignore or merged into this file. For a more nuclear +# option (not recommended) you can uncomment the following to ignore the entire idea folder. +#.idea/ diff --git a/QCQ_ResNet18.py b/QCQ_ResNet18.py new file mode 100644 index 0000000..fcae726 --- /dev/null +++ b/QCQ_ResNet18.py @@ -0,0 +1,62 @@ +import torch +from torch import nn +from torch.nn import functional as F + + +class ResBlock(nn.Module): + def __init__(self, ch_in, ch_out, stride=2): + super(ResBlock, self).__init__() + self.conv1 = nn.Conv2d(ch_in, ch_out, kernel_size=3, stride=stride, padding=1) # ! (h-3+2)/2 + 1 = h/2 图像尺寸减半 + self.bn1 = nn.BatchNorm2d(ch_out) + self.conv2 = nn.Conv2d(ch_out, ch_out, kernel_size=3, stride=1, padding=1) # ! h-3+2*1+1=h 图像尺寸没变化 + self.bn2 = nn.BatchNorm2d(ch_out) + + self.extra = nn.Sequential( + nn.Conv2d(ch_in, ch_out, kernel_size=1, stride=stride), # ! 这句话是针对原图像尺寸写的，要进行element wise add + # ! 因此图像尺寸也必须减半，(h-1)/2+1=h/2 图像尺寸减半 + nn.BatchNorm2d(ch_out) + ) + + def forward(self, x): + out = x + x = torch.relu(self.bn1(self.conv1(x))) + x = self.bn2(self.conv2(x)) + # shortcut + # ! element wise add [b,ch_in,h,w] [b,ch_out,h,w] 必须当ch_in = ch_out时才能进行相加 + out = x + self.extra(out) # todo self.extra强制把输出通道变成一致 + return out + + +class ResNet18(nn.Module): + def __init__(self): + super(ResNet18, self).__init__() + self.conv1 = nn.Sequential( + nn.Conv2d(3, 64, kernel_size=3, stride=1, padding=1), # ! 图像尺寸不变 + nn.BatchNorm2d(64) + ) + # 4个ResBlock + # [b,64,h,w] --> [b,128,h,w] + self.block1 = ResBlock(64, 128) + # [b,128,h,w] --> [b,256,h,w] + self.block2 = ResBlock(128, 256) + # [b,256,h,w] --> [b,512,h,w] + self.block3 = ResBlock(256, 512) + # [b,512,h,w] --> [b,512,h,w] + self.block4 = ResBlock(512, 512) + + self.outlayer = nn.Linear(512, 10) + + def forward(self, x): + x = torch.relu(self.conv1(x)) + # [b,64,h,w] --> [b,1024,h,w] + x = self.block1(x) + x = self.block2(x) + x = self.block3(x) + x = self.block4(x) + # print("after conv:",x.shape) + # [b,512,h,w] --> [b,512,1,1] + x = F.adaptive_avg_pool2d(x, [1, 1]) + # flatten + x = x.view(x.shape[0], -1) + x = self.outlayer(x) + return x diff --git a/QCQ_VGG16.py b/QCQ_VGG16.py new file mode 100644 index 0000000..53ed451 --- /dev/null +++ b/QCQ_VGG16.py @@ -0,0 +1,44 @@ + +from torch import nn + + +cfg = {'VGG16': [64, 64, 'M', 128, 128, 'M', 256, 256, 256, 'M', 512, 512, 512, 'M', 512, 512, 512, 'M']} +class QCQ(nn.Module): + + def __init__(self, vgg): + super(QCQ, self).__init__() + self.features = self._make_layers(vgg) + self.dense = nn.Sequential( + nn.Linear(512, 4096), + nn.ReLU(inplace=True), + nn.Dropout(0.4), + nn.Linear(4096, 4096), + nn.ReLU(inplace=True), + nn.Dropout(0.4), + ) + self.classifier = nn.Linear(4096, 10) + + def forward(self, x): + out = self.features(x) + out = out.view(out.size(0), -1) + out = self.dense(out) + out = self.classifier(out) + return out + + def _make_layers(self, vgg): + layers = [] + in_channels = 3 + for x in vgg: + if x == 'M': + layers += [nn.MaxPool2d(kernel_size=2, stride=2)] + else: + layers += [nn.Conv2d(in_channels, x, kernel_size=3, padding=1), + nn.BatchNorm2d(x), + nn.ReLU(inplace=True)] + in_channels = x + + layers += [nn.AvgPool2d(kernel_size=1, stride=1)] + return nn.Sequential(*layers) + + + diff --git a/README.md b/README.md index a461cd0..f87d3d0 100644 --- a/README.md +++ b/README.md @@ -1,2 +1,10 @@ # CIFAR10 +CIFAR10学习 +我的环境： +python 3.6.13 +torch 1.10.2 +cuda 11.3 +torchvision 0.11.3 +pillow 8.4.0 +tensorboard 2.10.1 \ No newline at end of file diff --git a/demo.py b/demo.py new file mode 100644 index 0000000..5c28b41 --- /dev/null +++ b/demo.py @@ -0,0 +1,97 @@ +import os + +import PIL +import torchvision.transforms +from PIL import Image + + +from QCQ_VGG16 import * +from QCQ_ResNet18 import * + +CIFAR10_class = ['airplane','automobile','brid','cat','deer','dog','frog','horse','ship','truck'] +vgg = [96, 96, 'M', 128, 128, 'M', 256, 256, 256, 'M', 512, 512, 512, 'M', 512, 512, 512, 'M'] + +def readImage(img_path='img/test.png'): + img = Image.open(img_path).convert('RGB') + transform= torchvision.transforms.Compose([torchvision.transforms.Resize((32,32)),torchvision.transforms.ToTensor()]) + img = transform(img) + print(img.shape) + img = torch.reshape(img, (1, 3, 32, 32)) + return img + + +def DrawImageTxt(imageFile, targetImageFile, txtnum): + # 设置字体大小 + font = PIL.ImageFont.truetype('img/abc.ttf', 50) + # 打开文件 + im = Image.open(imageFile) + # 字体坐标 + draw = PIL.ImageDraw.Draw(im) + draw.text((0, 0), txtnum, (255, 255, 0), font=font) + # 保存 + im.save(targetImageFile) + # 关闭 + im.close() + + +if __name__=='__main__': + print("通过训练好的模型识别十种事物：飞机,汽车,鸟,猫,鹿,狗,青蛙,马,船,卡车") + device = 'cuda' + models=[] + i=1 + for root, dirs, files in os.walk('model'): + for file in files: + if file.__contains__('.pth'): + file_path=root+'/'+file + models.append(file_path) + print(f'{i}.'+file) + i+=1 + if models.__len__()==0: + print('model文件夹中没有pth模型文件') + else: + select = int(input('选择一个模型\n')) + model_path=models[select-1] + if model_path.__contains__('VGG16'): + qcq_test=QCQ(vgg) + elif model_path.__contains__('RestNet18'): + qcq_test=ResNet18() + else: + print('选择的模型名称中既不包含"VGG16"，也不包含"RestNet18"') + + qcq_test.load_state_dict(torch.load(model_path, map_location=torch.device(device))) + + imgs=[] + i=1 + for root, dirs, files in os.walk('img'): + for file in files: + if file.__contains__('.png') or file.__contains__('.jpg'): + if not file.__contains__('pre'): + file_path = root + '/' + file + imgs.append(file_path) + print(f'{i}.' + file) + i+=1 + if imgs.__len__() == 0: + print('img文件夹中没有图片') + else: + select = int(input('选择一个测试图片,\n')) + image_path = imgs[select-1] + image_name,image_type = image_path.split('.') + targetImageFile=image_name+'_pre.png' + img=readImage(image_path) + qcq_test.eval() + with torch.no_grad(): + output = qcq_test(img) + pre=output.argmax(1) + txtnum = CIFAR10_class[pre.item()] + DrawImageTxt(image_path, targetImageFile, txtnum) + print(output) + print(pre) + print(txtnum) + Image.open(targetImageFile).show() + + + + + + + diff --git a/img/abc.ttf b/img/abc.ttf new file mode 100644 index 0000000..c5f12a8 Binary files /dev/null and b/img/abc.ttf differ diff --git a/img/test.png b/img/test.png new file mode 100644 index 0000000..c34f02b Binary files /dev/null and b/img/test.png differ diff --git a/img/test2.png b/img/test2.png new file mode 100644 index 0000000..5bfb099 Binary files /dev/null and b/img/test2.png differ diff --git a/img/test2_pre.png b/img/test2_pre.png new file mode 100644 index 0000000..cd19fb5 Binary files /dev/null and b/img/test2_pre.png differ diff --git a/img/test_pre.png b/img/test_pre.png new file mode 100644 index 0000000..c8116b2 Binary files /dev/null and b/img/test_pre.png differ diff --git a/train_CIFAR10_complete.py b/train_CIFAR10_complete.py new file mode 100644 index 0000000..6af27c0 --- /dev/null +++ b/train_CIFAR10_complete.py @@ -0,0 +1,124 @@ +import os.path + + +import torchvision.datasets +from torch.utils.tensorboard import SummaryWriter + +from QCQ_VGG16 import * +from QCQ_ResNet18 import * +from torch.utils.data import DataLoader + +# 训练数据集 +train_data = torchvision.datasets.CIFAR10("CIFAR10/CIFAR10_train", train=True, + transform=torchvision.transforms.ToTensor(), download=True) +# 测试数据集 +test_data = torchvision.datasets.CIFAR10("CIFAR10/CIFAR10_test", train=False, + transform=torchvision.transforms.ToTensor(), download=True) + +# 数据集长度 +train_data_size = len(train_data) +test_data_size = len(test_data) +print("训练数据集的长度为：{}".format(train_data_size)) +print("测试数据集的长度为：{}".format(test_data_size)) + +# dataloader加载数据集 +train_dataloader = DataLoader(train_data, batch_size=64) +test_dataloader = DataLoader(test_data, batch_size=64) + +# 定义运行设备 +device = torch.device('cuda' if torch.cuda.is_available() else 'cpu') + +# 创建神经网络模型 +vgg = [96, 96, 'M', 128, 128, 'M', 256, 256, 256, 'M', 512, 512, 512, 'M', 512, 512, 512, 'M'] + +select=input("选择模型:\n" + "1.VGG16\n" + "2.RestNet18\n") +if select=='1': + qcq = QCQ(vgg) + model_name='VGG16' + learning_rate = 1e-2 + optimizer = torch.optim.SGD(qcq.parameters(), lr=learning_rate,weight_decay=5e-3) + scheduler = torch.optim.lr_scheduler.StepLR(optimizer, step_size=5, gamma=0.4, last_epoch=-1) + # 训练轮数 + epoch = 10 +else: + qcq = ResNet18() + model_name = 'RestNet18' + # 优化器,学习率 + learning_rate = 1e-3 + optimizer = torch.optim.Adam(qcq.parameters(), lr=learning_rate) + # 训练轮数 + epoch = 50 +qcq = qcq.to(device) + +# 创建损失函数 +loss_fun = nn.CrossEntropyLoss().to(device) +loss_fun = loss_fun.to(device) + + + +# 设置训练网络的一些参数 +# 训练次数 +total_train_step = 0 +# 测试次数 +total_test_step = 0 + + +# 添加tensorboard +if not os.path.exists('logs'): + os.mkdir('logs') +writer = SummaryWriter('logs') + +for i in range(epoch): + print("--------第%5d 轮训练开始--------" % (i + 1)) + + # 训练步骤开始 + qcq.train() + for data in train_dataloader: + imgs, targets = data + + imgs = imgs.to(device) + targets = targets.to(device) + + outputs = qcq(imgs) + loss = loss_fun(outputs, targets) + # 模型优化 + optimizer.zero_grad() + loss.backward() + optimizer.step() + total_train_step += 1 + if total_train_step % 100 == 0: + print('训练次数：%-8d,Loss：%f' % (total_train_step, loss.item())) + writer.add_scalar('train_loss', loss.item(), total_train_step) + + # 测试步骤开始 + qcq.eval() + total_test_loss = 0 + total_accuracy = 0 + with torch.no_grad(): + for data in test_dataloader: + imgs, targets = data + + imgs = imgs.to(device) + targets = targets.to(device) + + outputs = qcq(imgs) + loss = loss_fun(outputs, targets) + total_test_loss += loss.item() + accuracy = (outputs.argmax(1) == targets).sum() + total_accuracy += accuracy + print('整体测试集上test_loss：%f' % total_test_loss) + print('整体测试集上正确率：%f' % (total_accuracy / test_data_size)) + writer.add_scalar('test_loss', total_test_loss, total_test_step) + writer.add_scalar('test_accuracy_rate', total_accuracy / test_data_size, total_test_step) + if select=='1': scheduler.step() + total_test_step += 1 + + # 模型每轮保存 + if not os.path.exists('model'): + os.mkdir('model') + torch.save(qcq.state_dict(), 'model/'+model_name+'_%d_epoch.pth' % (i+1)) + print("模型自动保存成功") + +writer.close()