How to process two images with the same augmentation?

You could use the functional API via torchvision.transforms.functional as seen in this example.