BEGIN:VCALENDAR VERSION:2.0 PRODID:researchseminars.org CALSCALE:GREGORIAN X-WR-CALNAME:researchseminars.org BEGIN:VEVENT SUMMARY:Carola-Bibiane Schönlieb (University of Cambridge) DTSTART:20210218T170000Z DTEND:20210218T180000Z DTSTAMP:20260423T021145Z UID:Inverse/34 DESCRIPTION:Title: Machine Learned Regularization for Solving Inverse Problems\nby Carol a-Bibiane Schönlieb (University of Cambridge) as part of International Zo om Inverse Problems Seminar\, UC Irvine\n\n\nAbstract\nInverse problems ar e about the reconstruction of an unknown physical quantity from indirect m easurements. Most inverse problems of interest are ill-posed and require a ppropriate mathematical treatment for recovering meaningful solutions. Reg ularization is one of the main mechanisms to turn inverse problems into we ll-posed ones by adding prior information about the unknown quantity to th e problem\, often in the form of assumed regularity of solutions. Classica lly\, such regularization approaches are handcrafted. Examples include Tik honov regularization\, the total variation and several sparsity-promoting regularizers such as the L1 norm of Wavelet coefficients of the solution. While such handcrafted approaches deliver mathematically and computational ly robust solutions to inverse problems\, providing a universal approach t o their solution\, they are also limited by our ability to model solution properties and to realise these regularization approaches computationally. Recently\, a new paradigm has been introduced to the regularization of in verse problems\, which derives regularization approaches for inverse probl ems in a data driven way. Here\, regularization is not mathematically mode lled in the classical sense\, but modelled by highly over-parametrised mod els\, typically deep neural networks\, that are adapted to the inverse pro blems at hand by appropriately selected (and usually plenty of) training d ata. In this talk\, I will review some machine learning based regularizati on techniques\, present some work on unsupervised and deeply learned conve x regularisers and their application to image reconstruction from tomograp hic and blurred measurements\, and finish by discussing some open mathemat ical problems.\n LOCATION:https://researchseminars.org/talk/Inverse/34/ END:VEVENT END:VCALENDAR