Weakly supervised learning with microscopy images and eggs

Hi,

I am trying to detect which parts of a microscopy image are out of focus and how far (from z-stacked images). Sadly I only have data that is not annotated. I have focus values for the main focus plain but some of the objects that should be in focus are not always on that plane. I also have some annotated data with these objects (possibly later on also annotations on which plane is in focus).

​

What do you …

deeplearning images learning supervised learning