Pytorch中的数据转换Transforms与DataLoader

DataLoader

DataLoader是一个比较重要的类，它为我们提供的常用操作有：batch_size(每个batch的大小), shuffle(是否进行shuffle操作), num_workers(加载数据的时候使用几个子进程)

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

import torch
'''
初始化网络
初始化Loss函数 & 优化器
进入step循环：
　　梯度清零
　　向前传播
　　计算本次Loss
　　向后传播
　　更新参数
'''
class LeNet(nn.Module):
    def __init__(self):
        super(LeNet, self).__init__()
        self.conv1 = nn.Conv2d(1, 10, kernel_size=5)
        self.conv2 = nn.Conv2d(10, 20, kernel_size=5)
        self.conv2_drop = nn.Dropout2d()
        self.fc1 = nn.Linear(320, 50)
        self.fc2 = nn.Linear(50, 10)
    def forward(self, x):
        x = F.relu(F.max_pool2d(self.conv1(x), 2))
        x = F.relu(F.max_pool2d(self.conv2_drop(self.conv2(x)), 2))
        x = x.view(-1, 320)
        x = F.relu(self.fc1(x))
        x = F.dropout(x, training=self.training)
        x = self.fc2(x)
        return x


if __name__ == "__main__":
    net = LeNet()

    # #########训练网络#########
    from torch import optim
    # from torchvision.datasets import MNIST
    import  torchvision
    import numpy
    from torchvision import transforms
    from torch.utils.data import DataLoader

    # 初始化Loss函数 & 优化器
    loss_fn = nn.CrossEntropyLoss()
    optimizer = optim.SGD(net.parameters(), lr=0.001, momentum=0.9)

    # transforms = transforms.Compose([])

    DOWNLOAD = False
    BATCH_SIZE = 32
    transform = transforms.Compose([
        transforms.ToTensor()
    ])
    #transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))  # 归一化

    train_dataset = torchvision.datasets.MNIST(root='./', train=True, transform=transform, download=DOWNLOAD)
    test_dataset = torchvision.datasets.MNIST(root='./data/mnist',
                                              train=False,
                                              transform=torchvision.transforms.ToTensor(),
                                              download=True)
   
    train_loader = DataLoader(dataset=train_dataset,
                              batch_size=BATCH_SIZE,
                              shuffle=True)
    test_loader = DataLoader(dataset=test_dataset, batch_size=BATCH_SIZE)

    for epoch in range(200):
        running_loss = 0.0
        for step, data in enumerate(train_loader):  
            inputs, labels = data
            inputs, labels = t.autograd.Variable(inputs), t.autograd.Variable(labels)
            # inputs = torch.from_numpy(inputs).unsqueeze(1)
            # labels = torch.from_numpy(numpy.array(labels))
            # 梯度清零
            optimizer.zero_grad()

            # forward
            outputs = net(inputs)
            # backward
            loss = loss_fn(outputs, labels)
            loss.backward()
            # update
            optimizer.step()

            running_loss += loss.item()
            if step % 10 == 9:
                print("[{0:d}, {1:5d}] loss: {2:3f}".format(epoch + 1, step + 1, running_loss / 2000))
                running_loss = 0.
    print("Finished Training")

   # save the trained net
    torch.save(net, 'net.pkl')

    # load the trained net
    net1 = torch.load('net.pkl')

    # test the trained net
    correct = 0
    total = 1
    for images, labels in test_loader:
        preds = net(images)
        predicted = torch.argmax(preds, 1)
        total += labels.size(0)
        correct += (predicted == labels).sum().item()

    accuracy = correct / total
    print('accuracy of test data:{:.1%}'.format(accuracy))

---

数据变换（Transform）

实例化数据库的时候，有一个可选的参数可以对数据进行转换，满足大多神经网络的要求输入固定尺寸的图片，因此要对原图进行Rescale或者Crop操作，然后返回的数据需要转换成Tensor。

数据转换（Transfrom）发生在数据库中的__getitem__操作中。

class Rescale(object):
    """Rescale the image in a sample to a given size.

    Args:
        output_size (tuple or int): Desired output size. If tuple, output is
            matched to output_size. If int, smaller of image edges is matched
            to output_size keeping aspect ratio the same.
    """

    def __init__(self, output_size):
        assert isinstance(output_size, (int, tuple))
        self.output_size = output_size

    def __call__(self, sample):
        image, landmarks = sample['image'], sample['landmarks']

        h, w = image.shape[:2]
        if isinstance(self.output_size, int):
            if h > w:
                new_h, new_w = self.output_size * h / w, self.output_size
            else:
                new_h, new_w = self.output_size, self.output_size * w / h
        else:
            new_h, new_w = self.output_size

        new_h, new_w = int(new_h), int(new_w)

        img = transform.resize(image, (new_h, new_w))

        # h and w are swapped for landmarks because for images,
        # x and y axes are axis 1 and 0 respectively
        landmarks = landmarks * [new_w / w, new_h / h]

        return {
    
    'image': img, 'landmarks': landmarks}


class RandomCrop(object):
    """Crop randomly the image in a sample.

    Args:
        output_size (tuple or int): Desired output size. If int, square crop
            is made.
    """

    def __init__(self, output_size):
        assert isinstance(output_size, (int, tuple))
        if isinstance(output_size, int):
            self.output_size = (output_size, output_size)
        else:
            assert len(output_size) == 2
            self.output_size = output_size

    def __call__(self, sample):
        image, landmarks = sample['image'], sample['landmarks']

        h, w = image.shape[:2]
        new_h, new_w = self.output_size

        top = np.random.randint(0, h - new_h)
        left = np.random.randint(0, w - new_w)

        image = image[top: top + new_h,
                      left: left + new_w]

        landmarks = landmarks - [left, top]

        return {
    
    'image': image, 'landmarks': landmarks}


class ToTensor(object):
    """Convert ndarrays in sample to Tensors."""

    def __call__(self, sample):
        image, landmarks = sample['image'], sample['landmarks']

        # swap color axis because
        # numpy image: H x W x C
        # torch image: C X H X W
        image = image.transpose((2, 0, 1))
        return {
    
    'image': torch.from_numpy(image),
                'landmarks': torch.from_numpy(landmarks)}

---

torchvision 包的介绍

torchvision 是PyTorch中专门用来处理图像的库，这个包中有四个大类。

torchvision.datasets
torchvision.models
torchvision.transforms
torchvision.utils

torchvision.datasets

torchvision.datasets 是用来进行数据加载的，PyTorch团队在这个包中帮我们提前处理好了很多很多图片数据集。

MNIST、COCO、Captions、Detection、LSUN、ImageFolder、Imagenet-12、CIFAR、STL10、SVHN、PhotoTour

import torchvision
from torch.utils.data import DataLoader

DOWNLOAD = False
BATCH_SIZE = 32
transform = transforms.Compose([
    transforms.ToTensor()
])
#transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))  # 归一化

train_dataset = torchvision.datasets.MNIST(root='./', train=True, transform=transform, download=DOWNLOAD)

train_loader = DataLoader(dataset=train_dataset,
                         batch_size=BATCH_SIZE,
                         shuffle=True)

torchvision.models

torchvision.models 中为我们提供了已经训练好的模型，加载之后，可以直接使用。包含以下模型结构。

AlexNet、VGG、ResNet、SqueezeNet、DenseNet、MobileNet

import torchvision.models as models
resnet18 = models.resnet18(pretrained=True)
alexnet = models.alexnet(pretrained=True)

torchvision.transforms

transforms提供了一般图像的转化操作类

# 图像预处理步骤
transform = transforms.Compose([
    transforms.Resize(96), # 缩放到 96 * 96 大小
    transforms.ToTensor(),
    transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5)) # 归一化
])

---

Transforms支持的变化

参考Pytorch中文文档

---

__all__ = ["Compose", "ToTensor", "PILToTensor", "ConvertImageDtype", "ToPILImage", "Normalize", "Resize", "Scale",
           "CenterCrop", "Pad", "Lambda", "RandomApply", "RandomChoice", "RandomOrder", "RandomCrop",
           "RandomHorizontalFlip", "RandomVerticalFlip", "RandomResizedCrop", "RandomSizedCrop", "FiveCrop", "TenCrop",
           "LinearTransformation", "ColorJitter", "RandomRotation", "RandomAffine", "Grayscale", "RandomGrayscale",
           "RandomPerspective", "RandomErasing", "GaussianBlur", "InterpolationMode", "RandomInvert", "RandomPosterize",
           "RandomSolarize", "RandomAdjustSharpness", "RandomAutocontrast", "RandomEqualize"]

---

from PIL import Image
# from torch.utils.tensorboard import SummaryWriter
from torchvision import transforms

from torch.autograd import Variable
from torchvision.transforms import functional as F

tensor数据类型
# 通过transforms.ToTensor去看两个问题

img_path = "./k.jpg"
img = Image.open(img_path)

# writer = SummaryWriter("logs")

tensor_trans = transforms.ToTensor()
tensor_img = tensor_trans(img)

tensor_img1 = F.to_tensor(img)

print(tensor_img.type(),tensor_img1.type())
print(tensor_img.shape)

'''
transforms.Normalize使用如下公式进行归一化：
channel=（channel-mean）/std(因为transforms.ToTensor()已经把数据处理成[0,1],那么(x-0.5)/0.5就是[-1.0, 1.0])
'''

# writer.add_image("Tensor_img", tensor_img)
# writer.close()

---

将输入的PIL.Image重新改变大小成给定的size，size是最小边的边长。举个例子，如果原图的height>width,那么改变大小后的图片大小是(size*height/width, size)。

### class torchvision.transforms.Scale(size, interpolation=2)

```python
from torchvision import transforms
from PIL import Image
crop = transforms.Scale(12)
img = Image.open('test.jpg')

print(type(img))
print(img.size)

croped_img=crop(img)
print(type(croped_img))
print(croped_img.size)

---

对PIL.Image进行变换

class torchvision.transforms.Compose(transforms)

将多个transform组合起来使用。

---

class torchvision.transforms.Normalize(mean, std)

给定均值：(R,G,B) 方差：（R，G，B），将会把Tensor正则化。即：Normalized_image=(image-mean)/std。

---

class torchvision.transforms.RandomSizedCrop(size, interpolation=2)

先将给定的PIL.Image随机切，然后再resize成给定的size大小。

---

class torchvision.transforms.RandomCrop(size, padding=0)

切割中心点的位置随机选取。size可以是tuple也可以是Integer。

---

class torchvision.transforms.CenterCrop(size)

将给定的PIL.Image进行中心切割，得到给定的size，size可以是tuple，(target_height, target_width)。size也可以是一个Integer，在这种情况下，切出来的图片的形状是正方形。