What is the scale of the sizes used for AnchorGenerator()?

Hello everyone,
I recently stumbled upon this snippet of code given below, but I couldnt find any information concerning the scales of these numbers. what are they? are they in pixels? or they are sth entirely different?

from torchvision.models.detection.faster_rcnn import FasterRCNN
from torchvision.models.detection.rpn import AnchorGenerator

anchors = AnchorGenerator(sizes=(32,64,128,256,512), 
                          aspect_ratios=(0.5,1.0,2.0))

Thanks a lot in advance