BEGIN:VCALENDAR
VERSION:2.0
PRODID:researchseminars.org
CALSCALE:GREGORIAN
X-WR-CALNAME:researchseminars.org
BEGIN:VEVENT
SUMMARY:Juyong Zhang
DTSTART:20201228T074500Z
DTEND:20201228T083000Z
DTSTAMP:20260423T024028Z
UID:iccm2020/62
DESCRIPTION:Title: <a href="https://researchseminars.org/talk/iccm2020/62/
 ">Anderson Acceleration for First-Order Methods in Visual Computing</a>\nb
 y Juyong Zhang as part of ICCM 2020\n\n\nAbstract\nAlternating optimizatio
 n methods like local-global solver and alternating direction multiplier me
 thod (ADMM) are commonly used in areas like signal processing\, machine le
 arning and computer graphics. These methods converge quickly to an approxi
 mate solution\, but can take a long time to converge to a solution of high
 -accuracy. In this talk\, I will present our works about applying Anderson
  acceleration to speed up the convergence of these methods by treating the
 m as fixed-point iteration. We also analyze the convergence of the propose
 d acceleration method on nonconvex problems\, and verify its effectiveness
  on a variety of problems.\n
LOCATION:https://researchseminars.org/talk/iccm2020/62/
END:VEVENT
END:VCALENDAR