ECE ILLINOIS Facebook ECE ILLINOIS on Twitter ECE ILLINOIS Alumni and Friends on LinkedIn ECE ILLINOIS Instagram

July 2014

  S M T W T F S
> 29 30 1 2 3 4 5
> 6 7 8 9 10 11 12
> 13 14 15 16 17 18 19
> 20 21 22 23 24 25 26
> 27 28 29 30 31 1 2

Advanced Search




Add this event to my calendar

“Compressed sensing of sparse rank-one matrices”

Speaker Kiryung Lee
Date: 1/27/2014
Time: 4:00 pm
Location: 141 Coordinated Science Lab
Event Contact: Denise Lewis
217-300-0222
edlewis@illinois.edu
Sponsor: Coordinated Science Lab
  ***Abstract*** Sparse bilinear inverse problems arise in numerous applications including medical imaging, blind deconvolution, and dictionary learning. Recasting these inverse problems as compressed sensing of sparse rank-one matrices, we propose a computationally efficient alternating minimization algorithm called sparse power factorization (SPF). Starting from a particular initialization, SPF achieves guaranteed stable recovery with number of measurements within a logarithmic factor of the information-theoretic fundamental limit. Numerical results show that SPF empirically outperforms the best known combinations of mixed norm and nuclear norm. Joint work with Yihong Wu and Yoram Bresler. ****Bio**** Kiryung Lee received the B.S. and M.S. degrees in electrical engineering and computer science from Seoul National University, Seoul, Korea, in 2000 and2002, respectively, and Ph.D. degree in electrical and computer engineering from the University of Illinois at Urbana-Champaign, in 2012. From 2002 to 2004, he was a member of engineering staff at the Electronic Telecommunications Research Institute, Daejeon, Korea. From 2004 to 2006, he was a research engineer at the LG Electronics, Seoul, Korea. He is currently a postdoctoral researcher at the University of Illinois at Urbana-Champaign. His current research interests include compressed sensing, inverse problems in imaging and signal processing, convex optimization, and noncommutative probability theory.