Michael Jordan, Professor of Electrical Engineering and Computer Science, Statistics, and co-collaborators, Amit Kohli, Stephen Bates, Jitendra Malik, Thayer Alshaabi, Srigokul Upadhyayula, and Yaniv Romano published a research paper addressing image-to-image regression by developing uncertainty quantification techniques. These techniques are designed to protect against a model’s mistakes and hallucinations. They used quantitative phase microscopy, accelerated magnetic resonance imaging, and super-resolution transmission electron microscopy of a Drosophila melanogaster brain to evaluate their methods.
Read more about it here.