BEGIN:VCALENDAR VERSION:2.0 PRODID:researchseminars.org CALSCALE:GREGORIAN X-WR-CALNAME:researchseminars.org BEGIN:VEVENT SUMMARY:Rebecca Willett (University of Chicago) DTSTART:20210507T130000Z DTEND:20210507T140000Z DTSTAMP:20260423T003253Z UID:MPML/41 DESCRIPTION:Title: Ma chine Learning and Inverse Problems: Deeper and More Robust\nby Rebecc a Willett (University of Chicago) as part of Mathematics\, Physics and Mac hine Learning (IST\, Lisbon)\n\n\nAbstract\nMany challenging image process ing tasks can be described by an ill-posed linear inverse problem: deblurr ing\, deconvolution\, inpainting\, compressed sensing\, and superresolutio n all lie in this framework. Recent advances in machine learning and image processing have illustrated that it is often possible to learn a regulari zer from training data that can outperform more traditional approaches by large margins. In this talk\, I will describe the central prevailing theme s of this emerging area and a taxonomy that can be used to categorize diff erent problems and reconstruction methods. We will also explore mechanisms for model adaptation\; that is\, given a network trained to solve an init ial inverse problem with a known forward model\, we propose novel procedur es that adapt the network to a perturbed forward model\, even without full knowledge of the perturbation. Finally\, I will describe a new class of a pproaches based on "infinite-depth networks" that can yield up to a 4dB PS NR improvement in reconstruction accuracy above state-of-the-art alternati ves and where the computational budget can be selected at test time to opt imize context-dependent trade-offs between accuracy and computation.\n LOCATION:https://researchseminars.org/talk/MPML/41/ END:VEVENT END:VCALENDAR