R440, Astronomy-Mathematics Building, NTU
(台灣大學天文數學館 440室)
A general framework for a class of first order primal-dual algorithms for convex optimization in imaging science
Yi-Su Lo (National Central University)
In this talk, we introduced an efficient prime-dual hybrid gradient algorithm for total variation minimization problems. This descent-type algorithm alternates between primal and dual variables and could be connected to several popular existing methods such as CGM and Chambolle’s algorithms. As a numerical experiment, we also present a comparison of the performance of these algorithms applied to a benchmark denoising problem in the image processing realm.
Reference
Esser, Ernie, Xiaoqun Zhang, and Tony F. Chan. "A general framework for a class of first order primal-dual algorithms for convex optimization in imaging science." SIAM Journal on Imaging Sciences 3.4 (2010): 1015-1046.
