Pytorch load data

One 、 Load data

• Load a single image in the dataset

from PIL import Image

img_path="D:\\pycharm\\PycharmProjects\\learn_torch\\hymenoptera_data\\train\\ants\\0013035.jpg"
img = Image.open(img_path)
img.show()

• Load dataset list  

import os

dir_path = "hymenoptera_data/train/ants"
img_path_list = os.listdir(dir_path)

stay pycharm Medium Python console（ controller ） Run in , The effect is more intuitive .

Two 、 Application pytorch Load data

torch.utils  yes torch Common toolbox .

Want to according to  idx  Get the corresponding picture , Get the picture address first list( obtain dataset All under data).

from torch.utils.data import Dataset

•   Definition  MyData  class

class MyData(Dataset):

    ''' Define global variables '''
    def __init__(self, root_dir, label_dir):
        self.root_dir = root_dir  # root_dir  Is the directory of the dataset 
        self.label_dir = label_dir  # label_dir  Is the catalog of labels 
        self.path = os.path.join(self.root_dir, self.label_dir)
        self.img_path = os.listdir(self.path)
 
    ''' get data '''
    def __getitem__(self, idx):
        img_name = self.img_path[idx]
        img_item_path = os.path.join(self.root_dir, self.label_dir, img_name)
        img = Image.open(img_item_path)
        label = self.label_dir
        return img, label

    ''' The length of the data set '''
    def __len__(self):
        return len(self.img_path)

•   Test defined  MyData  class

root_dir = "hymenoptera_data/train"
ants_label_dir = "ants"
bees_label_dir = "bees"
ants_dataset = MyData(root_dir, ants_label_dir)
bees_dataset = MyData(root_dir, bees_label_dir)

Method 1 ： 

Can be found in Python Medium test ： 

ants_dataset[0]

  Method 2 ：

img, label = ants_dataset[0]
img.show()

  Add ：

train_dataset = ants_dataset + bees_dataset  #  The whole training set （ A collection of ant data sets and bee data sets ）

When the data set is insufficient , Data sets can be supplemented in this way