Pytorch入门学习（六）--- 加载数据以及预处理（初步）--- 只为了更好理解流程

直接从Pytorch Tutorials拿过来，看看。

需要的包:
1. scikit-image: 图像io以及变形
2. pandas: 读入csv文件
数据：
faces
csv的数据形式:
总共68个人脸关键点。

     image_name,part_0_x,part_0_y,part_1_x,part_1_y,part_2_x, ... ,part_67_x,part_67_y     0805personali01.jpg,27,83,27,98, ... 84,134     1084239450_e76e00b7e7.jpg,70,236,71,257, ... ,128,312

或是如图:

最简单的通过函数读取图片

# -*- coding: utf-8 -*-"""Data Loading and Processing Tutorial====================================**Author**: `Sasank Chilamkurthy <https://chsasank.github.io>`_"""from __future__ import print_function, divisionimport osimport torchimport pandas as pdfrom skimage import io, transformimport numpy as npimport matplotlib.pyplot as pltfrom torch.utils.data import Dataset, DataLoaderfrom torchvision import transforms, utils# Ignore warningsimport warningswarnings.filterwarnings("ignore")plt.ion()   # interactive modelandmarks_frame = pd.read_csv('faces/face_landmarks.csv')n = 65img_name = landmarks_frame.ix[n, 0]landmarks = landmarks_frame.ix[n, 1:].as_matrix().astype('float')landmarks = landmarks.reshape(-1, 2)print('Image name: {}'.format(img_name))print('Landmarks shape: {}'.format(landmarks.shape))print('First 4 Landmarks: {}'.format(landmarks[:4]))# 定义show_landmarks.def show_landmarks(image, landmarks):    """Show image with landmarks"""    plt.imshow(image)    plt.scatter(landmarks[:, 0], landmarks[:, 1], s=10, marker='.', c='r')    plt.pause(0.001)  # pause a bit so that plots are updatedplt.figure()# 用 io.imread来读取图片show_landmarks(io.imread(os.path.join('faces/', img_name)),               landmarks)plt.show()

通过继承Dataset

# 自定义数据集时，要继承 Dataset类。# 一般至少要有 __init__, __len__, __getitem__class FaceLandmarksDataset(Dataset):    """Face Landmarks dataset."""    def __init__(self, csv_file, root_dir, transform=None):        """        Args:            csv_file (string): Path to the csv file with annotations.            root_dir (string): Directory with all the images.            transform (callable, optional): Optional transform to be applied                on a sample.        """        self.landmarks_frame = pd.read_csv(csv_file)        self.root_dir = root_dir        self.transform = transform    def __len__(self):        return len(self.landmarks_frame)    def __getitem__(self, idx):        img_name = os.path.join(self.root_dir, self.landmarks_frame.ix[idx, 0])        image = io.imread(img_name)        landmarks = self.landmarks_frame.ix[idx, 1:].as_matrix().astype('float')        landmarks = landmarks.reshape(-1, 2)        sample = {'image': image, 'landmarks': landmarks}        if self.transform:            sample = self.transform(sample)  # 采用这个，不错。        return sampleface_dataset = FaceLandmarksDataset(csv_file='faces/face_landmarks.csv',                                    root_dir='faces/')fig = plt.figure()for i in range(len(face_dataset)):    sample = face_dataset[i]    print(i, sample['image'].shape, sample['landmarks'].shape)  # 对于dict，可以使用['key']来获取值。    ax = plt.subplot(1, 4, i + 1)    plt.tight_layout()    ax.set_title('Sample #{}'.format(i))    ax.axis('off')    show_landmarks(**sample)  # 关键字参数传参    if i == 3:        plt.show()        break

第三部分：图像变换的基本写法(继承object)

主要有：
1. 放缩
2. 随机裁剪
3. 将numpy 的图像数据转换为 tensor.

其他：
1. 这里将类变成可调用对象。通过类中实现 call()
比如：
tsfm = Transform( init_params)
transformed_sample = tsfm(call_params)
就是　类的实例(参数)。

知识点：
1. isinstance(output_size, (int, tuple))**后面的参数可以是tuple的。以此来同时验证多种类型。
2. numpy和pytorch的图像要进行swap axis.

        # 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)}

class Rescale(object):    """Rescale the image in a sample to a given size.    Args:        output_size (tuple or tuple): 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))  # 不错。        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)}

第四部分：利用torchvision来组合图像变换

如果我们想对图像进行多个变换，那么可以用 torchvision.transforms.Compose.

"""注意: 1. FaceLandmarksDataset是在第二部分我们自定义的类（继承自Dataset)。2. transforms.Compose可以将多个操作组合。3. 每个操作都是一个类，都有 __call__函数。__call__函数的参数是sample。我们只需要 MyTransforms(init_params)即可。"""transformed_dataset = FaceLandmarksDataset(csv_file='faces/face_landmarks.csv',                                           root_dir='faces/',                                           transform=transforms.Compose([                                               Rescale(256),                                               RandomCrop(224),                                               ToTensor() # to tensor初始化函数没有参数                                           ]))# 很ugly的加载数据集for i in range(len(transformed_dataset)):    sample = transformed_dataset[i]    print(i, sample['image'].size(), sample['landmarks'].size())    if i == 3:        break

通过这种for循环处理数据集的方法有很多问题：
1. 没有将数据以成批的形式处理
2. 没有shuffle数据
3. 没有并行加载数据，用multiprocessing workers.

第五部分：使用DataLoader

torch.utils.data.DataLoader是一个迭代器，具有上面所有功能特性。
方式

dataloader = DataLoader(经过各种变换的自定义的数据集, batch_size=4,                        shuffle=True, num_workers=4)

#定义dataloaddataloader = DataLoader(transformed_dataset, batch_size=4,                        shuffle=True, num_workers=4)# Helper function to show a batchdef show_landmarks_batch(sample_batched):    """Show image with landmarks for a batch of samples."""    images_batch, landmarks_batch = \            sample_batched['image'], sample_batched['landmarks']    batch_size = len(images_batch)    im_size = images_batch.size(2)    grid = utils.make_grid(images_batch)    plt.imshow(grid.numpy().transpose((1, 2, 0)))    for i in range(batch_size):        plt.scatter(landmarks_batch[i, :, 0].numpy() + i * im_size,                    landmarks_batch[i, :, 1].numpy(),                    s=10, marker='.', c='r')        plt.title('Batch from dataloader')for i_batch, sample_batched in enumerate(dataloader):    print(i_batch, sample_batched['image'].size(),          sample_batched['landmarks'].size())    # observe 4th batch and stop.    if i_batch == 3:        plt.figure()        show_landmarks_batch(sample_batched)        plt.axis('off')        plt.ioff()        plt.show()        break

总结

综上有：

# -*- coding: utf-8 -*-############# 第一步： 准备 ##################"""Data Loading and Processing Tutorial====================================**Author**: `Sasank Chilamkurthy <https://chsasank.github.io>`_"""from __future__ import print_function, divisionimport osimport torchimport pandas as pdfrom skimage import io, transformimport numpy as npimport matplotlib.pyplot as pltfrom torch.utils.data import Dataset, DataLoaderfrom torchvision import transforms, utils# Ignore warningsimport warningswarnings.filterwarnings("ignore")plt.ion()   # interactive modelandmarks_frame = pd.read_csv('faces/face_landmarks.csv')n = 65img_name = landmarks_frame.ix[n, 0]landmarks = landmarks_frame.ix[n, 1:].as_matrix().astype('float')landmarks = landmarks.reshape(-1, 2)print('Image name: {}'.format(img_name))print('Landmarks shape: {}'.format(landmarks.shape))print('First 4 Landmarks: {}'.format(landmarks[:4]))# 定义show_landmarks.def show_landmarks(image, landmarks):    """Show image with landmarks"""    plt.imshow(image)    plt.scatter(landmarks[:, 0], landmarks[:, 1], s=10, marker='.', c='r')    plt.pause(0.001)  # pause a bit so that plots are updatedplt.figure()# 用 io.imread来读取图片show_landmarks(io.imread(os.path.join('faces/', img_name)),               landmarks)plt.show()################## 第二步：自定义dataset类 ####################class FaceLandmarksDataset(Dataset):    """Face Landmarks dataset."""    def __init__(self, csv_file, root_dir, transform=None):        """        Args:            csv_file (string): Path to the csv file with annotations.            root_dir (string): Directory with all the images.            transform (callable, optional): Optional transform to be applied                on a sample.        """        self.landmarks_frame = pd.read_csv(csv_file)        self.root_dir = root_dir        self.transform = transform    def __len__(self):        return len(self.landmarks_frame)    def __getitem__(self, idx):        img_name = os.path.join(self.root_dir, self.landmarks_frame.ix[idx, 0])        image = io.imread(img_name)        landmarks = self.landmarks_frame.ix[idx, 1:].as_matrix().astype('float')        landmarks = landmarks.reshape(-1, 2)        sample = {'image': image, 'landmarks': landmarks}        if self.transform:            sample = self.transform(sample)        return sample############# 第三步：自定义transforms类 ###################### 注意：自定义操作类，都扩展了 __call__函数，以便在后面 transforms.Compose调用。# 它们都继承object即可。class Rescale(object):    """Rescale the image in a sample to a given size.    Args:        output_size (tuple or tuple): 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)}############第四步: 数据集实例（自定义dataset类 + 各种变换组合）#############transformed_dataset = FaceLandmarksDataset(csv_file='faces/face_landmarks.csv',                                           root_dir='faces/',                                           transform=transforms.Compose([                                               Rescale(256),                                               RandomCrop(224),                                               ToTensor()                                           ]))############# 第五步: 使用DataLoader并行加载 ###################dataloader = DataLoader(transformed_dataset, batch_size=4,                        shuffle=True, num_workers=4)# Helper function to show a batchdef show_landmarks_batch(sample_batched):    """Show image with landmarks for a batch of samples."""    images_batch, landmarks_batch = \            sample_batched['image'], sample_batched['landmarks']    batch_size = len(images_batch)    im_size = images_batch.size(2)    grid = utils.make_grid(images_batch)    plt.imshow(grid.numpy().transpose((1, 2, 0)))    for i in range(batch_size):        plt.scatter(landmarks_batch[i, :, 0].numpy() + i * im_size,                    landmarks_batch[i, :, 1].numpy(),                    s=10, marker='.', c='r')        plt.title('Batch from dataloader')# 用dataLoader类就是一个batch的形式了。for i_batch, sample_batched in enumerate(dataloader):    print(i_batch, sample_batched['image'].size(),          sample_batched['landmarks'].size())    # observe 4th batch and stop.    if i_batch == 3:        plt.figure()        show_landmarks_batch(sample_batched)        plt.axis('off')        plt.ioff()        plt.show()        break

可以看到，这种方法需要自己写 各种transforms类，比较繁琐。但能很清晰地展示整个流程。下篇会介绍在真正项目中用到的方法。