Deep Learning Microscopy

  • I've taken a default stance with these kinds of applications that whatever extra detail we get is being "hallucinated" by the network. With something like microscopy, there could be serious consequences to such hallucinations. Wondering whether techniques exist to limit such effects.