Pytorch pad to square

import torchvision.transforms.functional as F

class SquarePad:
    def __call__(self, image):
        max_wh = max(image.size)
        p_left, p_top = [(max_wh - s) // 2 for s in image.size]
        p_right, p_bottom = [max_wh - (s+pad) for s, pad in zip(image.size, [p_left, p_top])]
        padding = (p_left, p_top, p_right, p_bottom)
        return F.pad(image, padding, 0, 'constant')

target_image_size = (224, 224)  # as an example
# now use it as the replacement of transforms.Pad class
transform=transforms.Compose([
    SquarePad(),
    transforms.Resize(target_image_size),
    transforms.ToTensor(),
    transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5)),
])

Source: discuss.pytorch.org

Similar codes

75 votes277 views6 answers
How to reverse string in python
Python
0 vote196 views1 answer
Python yeild
Python
0 vote163 views1 answer
Numpy compute mad
Python
2 votes197 views1 answer
Sqlalchemy query bilter by current month
Python
1 vote252 views1 answer
Python seaborn heatmap decrease annot size
Python
16 votes168 views4 answers
Tqdm library
Python
0 vote240 views2 answers
BUILD FAILED (Ubuntu 20.04 using python-build 20180424)
Python
0 vote224 views1 answer
Invalid HTTP_HOST header: 'dtodo2.herokuapp.com'. You may need to add 'dtodo2.herokuapp.com' to ALLOWED_HOSTS.
Python
0 vote206 views1 answer
Convert from 12 hrs to 24 python
Python
0 vote189 views3 answers
Find rules of decision tree python
Python