Cite Details
Paul Rodríguez and Brendt Wohlberg, "A Comparison of the Computational Performance of
Iteratively Reweighted Least Squares and Alternating Minimization
Algorithms for
ℓ1 Inverse Problems", in
Proceedings of IEEE International Conference on
Image Processing (ICIP), (Orlando, FL, USA), doi:
10.1109/ICIP.2012.6467548, pp. 3069--3072, Oct 2012
Abstract
Alternating minimization algorithms with a shrinkage step,
derived within the Split Bregman (SB) or Alternating Direction
Method of Multipliers (ADMM) frameworks, have become very popular
for ℓ1-regularized problems,
including Total Variation and Basis Pursuit Denoising. It appears to
be generally assumed that they deliver much better computational
performance than older methods such as Iteratively Reweighted Least
Squares (IRLS). We show, however, that IRLS type methods are
computationally competitive with SB/ADMM methods for a variety of
problems, and in some cases outperform them.
BibTeX Entry
@inproceedings{rodriguez-2012-comparison,
author = {Paul Rodr\'{i}guez and Brendt Wohlberg},
title = {A Comparison of the Computational Performance of
Iteratively Reweighted Least Squares and Alternating Minimization
Algorithms for
$\ell_{1}$
Inverse Problems},
year = {2012},
month = Oct,
urlpdf = {http://brendt.wohlberg.net/publications/pdf/rodriguez-2012-comparison.pdf},
booktitle = {Proceedings of IEEE International Conference on
Image Processing (ICIP)},
address = {Orlando, FL, USA},
doi = {10.1109/ICIP.2012.6467548},
pages = {3069--3072}
}
