Sina Jafarpour
PhD candidate,  Princeton University

Thursday, November 12, 2009
Schiciano Auditorium B
11:45 - 1:00pm

"Deterministic Compressive Sensing"

Compressed Sensing is a novel data acquisition method suitable when either the measurement or the compression is costly. In this talk, we present two efficient deterministic sensing methods. We first show how expander graphs are appropriate for compressed sensing in terms of providing explicit matrices as well as a simple and efficient recovery algorithm.  We then provide simple conditions a matrix should satisfy in order to satisfy statistical restricted isometry property. This leads to deterministic sensing matrices constructed from Kerdock codes and second order Reed-Muller codes, and a Chirp reconstruction algorithm for which the recovery time only depends on the number of measurements and may act better in the presence of noise.

Joint work with Robert Calderbank, Stephen Howard, Babak Hassibi and Weiyu Xu

Sina Jafarpour is a PhD candidate in Princeton University. He received his B.Sc in computer engineering from Sharif University of Technology in 2007, and is co-advised by Prof. Robert Calderbank, and Prof. Robert Schapire. His main research interests include compressive sensing, learning theory and boosting, and applications of machine learning in signal processing and information retrieval. Mr Jafarpour has been a member of the Van Gogh project supervised by Prof. Ingrid Daubechies, and the  Microsoft Chat-bot project supervised by Dr. Chris Burges.