2 p.m. - 3 p.m. Location: FN 2.102
Mathematical Sciences, U.T. Dallas
On recoverability properties of fixed measurement matrices in Compressed Sensing
Compressed Sensing (CS) is a technique that allows efficient compression and decompression of sparse vectors. There are two approaches to CS, probabilistic and deterministic. The probabilistic approach to CS is based upon certain types of random measurement matrices and the deterministic approach to CS is based upon construction of measurement matrices with the so called low coherence.
While both methods have their advantages and disadvantages, they both fail to answer the following question: Given an m by n matrix A with m<n, how sparse a vector U should be in order to be compressed and decompressed by matrix A.
In this talk I will answer this question by showing that recoverability properties of any measurement matrix can be computed in polynomial time complexity. In addition, by defining a dual of coherence I will prove that it is a better measure of recoverability properties of any matrix then currently used coherence.
Sponsored by the Department of Mathematical Sciences
John Zweck, 972-883-6699
Questions? Email me.