How to work with different shape images

vision
1

Hi,

I have a dataset with different image sizes.

Normally, I’m resizing the images to one size, but I can’t do it right now, since it’s an object detection problem.

How can I do it with pyTorch?