Many patch level tasks start with gigantic images, like 1536x2080 or something. Those images are then just naively resized down to 224x224 and processed.
This task is for coming up with a different method of handling these patches.
Preliminary thought is that we would chop the patch up and handle each section individually, then find a way to aggregate the results.