PyTorch读取Cifar数据集并显示图片的实例讲解

from torchvision import transforms, datasets as ds
from torch.utils.data import DataLoader
import matplotlib.pyplot as plt
import numpy as np

#transform = transforms.Compose是把一系列图片操作组合起来，比如减去像素均值等。
#DataLoader读入的数据类型是PIL.Image
#这里对图片不做任何处理，仅仅是把PIL.Image转换为torch.FloatTensor，从而可以被pytorch计算
transform = transforms.Compose(
 [
 transforms.ToTensor()
 ]
)

train_set = ds.CIFAR100(root='.', train=True, transform=transform, target_transform=None, download=True)

data_loader = DataLoader(dataset=train_set,
  batch_size=1,
  shuffle=False,
  num_workers=2)


# # 生成torch.utils.data.DataLoaderIter
# # 不过DataLoaderIter它会被DataLoader自动创建并且调用，我们用不到
# data_iter = iter(data_loader)
# images, labels = next(data_iter)

to_pil_image = transforms.ToPILImage()
cnt = 0
for image,label in data_loader:
 if cnt>=3: # 只显示3张图片
 break
 print(label) # 显示label

 # 方法1：Image.show()
 # transforms.ToPILImage()中有一句
 # npimg = np.transpose(pic.numpy(), (1, 2, 0))
 # 因此pic只能是3-D Tensor，所以要用image[0]消去batch那一维
 img = to_pil_image(image[0])
 img.show()

 # 方法2：plt.imshow(ndarray)
 img = image[0] # plt.imshow()只能接受3-D Tensor，所以也要用image[0]消去batch那一维
 img = img.numpy() # FloatTensor转为ndarray
 img = np.transpose(img, (1,2,0)) # 把channel那一维放到最后

 # 显示图片
 plt.imshow(img)
 plt.show()

 cnt += 1