BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Flower: A Flow Matching Solver for Inverse Problems
DTSTART:20260820T093000Z
DTEND:20260820T110000Z
DTSTAMP:20260815T184700Z
UID:indico-event-19250@indico.psi.ch
CONTACT:benjamin.bejar@psi.ch
DESCRIPTION:Speakers: Mehrsa Pourya (EPFL)\n\nAbstract: Inverse problems i
 nvolve the reconstruction of a signal from incomplete or noisy measurement
 s. Image restoration tasks\, such as super-resolution\, deblurring\, and i
 npainting\, are inverse problems. Their solution needs a prior on what nat
 ural images look like. Flow matching models are the current state of the a
 rt for image generation and provide exactly the kind of prior required. In
  this talk\, I will present Flower\, a method that transforms a pretrained
  flow matching model into a solver for inverse problems by modifying its i
 nference path. I will outline how Flower works\, show results for differen
 t tasks\, and discuss future work.\nTEAMS link here \n\nhttps://indico.psi
 .ch/event/19250/
LOCATION:OHSA/B17
URL:https://indico.psi.ch/event/19250/
END:VEVENT
END:VCALENDAR