| Lecturer Profile |
Trac D. Tran is Professor of the Depart. of ECE, Johns Hopkins Univ., MD. He is an IEEE Fellow. Dr. Tran received B.S. and M.S. from the MIT, USA, in 1993 and 1994, respectively, and Ph.D. from the Univ. of Wisconsin, Madison, in 1998, all in EE. His research interests are DSP, particularly in sparse representation, sparse recovery, sampling, multi-rate systems, filter banks, wavelets. He received the NSF CAREER award in 2001, and the Capers and Marion McDonald Award for Excellence in Mentoring and Advising in 2009. |
| Lecture Abstract |
This talk presents new theoretical ideas and mathematical frameworks on structured sparse data representations in the presence of low-rank interference. We provide various novel models for the representation of multiple measurements where signals exhibit a high level of correlation over a proper dictionary but are severely corrupted by certain interference which can be signals from external sources, underlying background, or any pattern noise that remains stationary during signal collection. Under the assumption that the interference component forms a low-rank structure, the proposed models minimize the rank of the interference to exclude it from the representation of multivariate sparse representation. We demonstrate that this is particularly essential in a multi-sensor problem since sensor co-location ensures that interference noise patterns are very similar, hence justifying the low-rank assumption. |
