Speaker
Jakob Knollmüller
(ORIGINS Data Science Lab)
Description
We introduce the Resolve method, a Bayesian approach for radio-interferometry that employs Variational Inference and Gaussian Processes. This method enables robust uncertainty quantification and advanced multi-domain imaging, essential for black hole studies. We demonstrate Resolve's effectiveness by confirming the Event Horizon Telescope (EHT) Collaboration's results on the supermassive black hole M87*. The method's performance in the next-generation EHT (ngEHT) Analysis challenge highlights its potential for future black hole research and dynamic reconstructions.
Primary author
Jakob Knollmüller
(ORIGINS Data Science Lab)
