How to load image using DataLoader with a list of image file names?

yttdebaba2 · June 20, 2017, 8:47am

I have a list containing a set of image file names and their corresponding labels.
These images are stored in a folder with some other images.
How can I load these images with DataLoader()?

Jake · June 20, 2017, 1:40pm

torchvision.datasets.ImageFolder(root, transform=None, target_transform=None, loader=)。this maybe help you.

yttdebaba2 · June 21, 2017, 3:49am

May I ask what is inside ‘loader’?
Is it a list class containing a set of image file names represented by str class?

chihyaoma · June 21, 2017, 6:05pm

I am not sure what you meant, but loader is a function to load files/images.

make_dataset() is where you define the list of files you are going to randomly select and pass to loader. By checking the code inside make_dataset(), you will get a better idea of what “list of files” will be collected.

github.com

pytorch/vision/blob/master/torchvision/datasets/folder.py#L24


    filename (string): path to a file


Returns:
    bool: True if the filename ends with a known image extension
"""
filename_lower = filename.lower()
return any(filename_lower.endswith(ext) for ext in IMG_EXTENSIONS)




def find_classes(dir):
classes = [d for d in os.listdir(dir) if os.path.isdir(os.path.join(dir, d))]
classes.sort()
class_to_idx = {classes[i]: i for i in range(len(classes))}
return classes, class_to_idx




def make_dataset(dir, class_to_idx):
images = []
dir = os.path.expanduser(dir)
for target in sorted(os.listdir(dir)):
    d = os.path.join(dir, target)