J. Zhu,
P. Han,
R. Niu,
A. Beirami, and
D. Baron,
"Large-Scale Multi-Processor Approximate Message Passing
with Lossy Compression,"
presented at Inf. Theory Applications Workshop,
San Diego, CA, February 2016
(slides).
P. Han,
J. Zhu,
R. Niu, and
D. Baron,
"Multi-Processor Approximate Message Passing Using Lossy Compression,"
Proc. Int. Conf. Acoustics, Speech, and Signal Process. (ICASSP2016),
Shanghai, China, March 2016
(pdf,
arxiv).
J. Zhu and
D. Baron,
"Multi-Processor Approximate Message Passing with Lossy Compression,"
submitted, January 2016
(pdf,
arxiv).
Teaching
ECE 421
(introduction to signal processing), January 2016.
J. Tan,
Y. Ma,
H. Rueda,
D. Baron, and
G. Arce,
"Approximate Message Passing in Coded Aperture Snapshot Spectral Imaging,"
Proc. IEEE Global Conf. Signal Inf. Process., Orlando, FL, Dec. 2015
(pdf,
talk,
arxiv).
J. Tan,
Y. Ma,
H. Rueda,
D. Baron, and
G. Arce,
"Compressive Hyperspectral Imaging via Approximate Message Passing,"
to appear in IEEE J. Sel. Topics Signal Process, March 2016.
(pdf,
arxiv,
tutorial video).
Congratulations to
Jin Tan,
for defending her Ph.D. dissertation, September 2015.
Teaching
ECE 308
(elements of control systems), August 2015.
R. Fayez, J. Young,
J. Tan,
Y. Ma, and
D. Baron,
"Image Reconstriction in Radio Astronomy,"
Duke Workshop on Sensing and Analysis of High-Dimensional Data,
Durham, NC, July 2015
(poster; no paper).
Y. Ma,
J. Zhu, and
D. Baron,
"Universal Denoising in Approximate Message Passing,"
Duke Workshop on Sensing and Analysis of High-Dimensional Data,
Durham, NC, July 2015
(poster; no paper).
Tutorial video by
Jin Tan and
Yanting Ma
about our research on compressive imaging and approximate
message passing. The video surveys two algorithmic works.
The first reconstructs a message measured noisily with linear
measurements; applications include medical imaging and radio astronomy.
The second reconstructs a hyper spectral cube acquired noisily
by a compressive hyper spectral system, for example the well-known CASSI system
(July 2015).
Tutorial video by
Yanting Ma and
Junan Zhu
about our research on universal denoising and approximate
message passing. This algorithm solves linear inverse problems
in a universal way without knowing the input statistics. (July 2015).
Tutorial video by
Junan Zhu
about our research on size- and level- adaptive
Markov chain Monte Carlo, which is an algorithm
that solves linear inverse problems in a universal
way without knowing the input statistics. (July 2015).
Tutorial video by
Nikhil Krishnan
about our research on parallel algorithms for universal compression
(July 2015).
Y. Ma, D.
Baron, and A. Beirami,
"Mismatched Estimation in Large Linear Systems,"
Proc. IEEE Int. Symp. Inf. Theory, Hong Kong, June 2015
(pdf,
talk,
arxiv).
H. J. Trussell and
D. Baron
"Creating Analytic Online Homework for Digital
Signal Processing,"
IEEE Signal Proc. Mag., vol. 32, no. 5,
pp. 112-118, Sept. 2015
(pdf).
N. Krishnan and
D. Baron,
"A Universal Parallel Two-Pass MDL
Context Tree Compression Algorithm,"
IEEE J. Sel. Topics Signal Process.,
vol. 9, no. 4, pp. 1-8, June 2015
(pdf,
arxiv,
tutorial video).
J. Tan,
Y. Ma, and
D. Baron,
"Compressive Imaging via Approximate
Message Passing with Image Denoising,"
IEEE Trans. Signal Proc., vol. 63, no. 8, pp. 2085-2092, April 2015
(pdf,
arxiv,
tutorial video).
J. Zhu,
D. Baron, and
M. F. Duarte,
"Recovery from Linear Measurements with Complexity-Matching
Universal Signal Estimation,"
IEEE Trans. Signal Proc., vol. 63, no. 6, pp. 1512-1527, March 2015
(arxiv,
pdf,
Matlab,
tutorial video).
Joe Young,
an unergraduate student working with our group,
will be joining Rice University's Ph.D. program. Good luck Joe!
February 2015.
Y. Ma,
J. Zhu, and
D. Baron,
"Universal Denoising and Approximate Message Passing,"
presented at Inf. Theory Applications Workshop,
San Diego, CA, February 2015
(talk).
Teaching
ECE 421
(introduction to signal processing), January 2015.
Research
During the last several years, we have been inundated by a deluge of data
in applications including distributed networked systems, finance, medical imaging, and seismics.
My interest lies in fundamental research for problems
involving vast amounts of data that must be processed effectively
and rapidly in order to extract useful - potentially "actionable" - information.
To approach these problems, we must use a
multi-disciplinary approach, and I combine tools
from information theory,
statistical signal processing, machine learning, and
computer science.
I call this computational information processing.
Specific research directions that I have worked on include:
In addition to teaching,
Joel Trussell
and I developed software for automating questions in
ECE 421
(Introduction to Signal Processing; undergraduate course)
using
WeBWorK
software. Each student receives a customized version of
each of the questions, and the student is allowed several
attempts to solve the question. The student may also request
another version of the question (with different numbers).
We used these for homeworks and quizzes during the 2015
spring semester.
Students solved the quizzes in class using laptops, tablets,
or even smart-phones; they received quiz grades immediately.
Overall, students provided favorable feedback
about the
WeBWorK-based
system, especially because it
allowed many small homeworks sets followed by brief quizzes,
which forced them to study consistently throughout the semester.
You are invited to check out some examples on our
demo
using a guest login. To learn more, please take a look at the
paper below. We would be glad to hear from you.
H. J. Trussell and
D. Baron
"Creating Analytic Online Homework for Digital
Signal Processing,"
to appear in IEEE Signal Proc. Mag., Sept. 2015
(pdf).
Joe Young
(B.S. 2015; currently Ph.D. student at Rice University).
Danielle Carmon (B.S. 2011; M.S. nonthesis 2012; currently with IBM).
Ilya Poltorak
(B.Sc. 2011 at
Technion;
currently a graduate student at Tel Aviv University).
Prospective Students
Regretfully, I cannot respond to most inquiries regarding openings for
graduate and postdoctoral positions in my group.
To get my attention, I suggest that you browse through my
webpage, see whether some of the research directions seem
interesting, and explain how it caught your attention.
I will very likely respond to such inquiries.
In contrast, prospective students who send the same letter to dozens or even
hundreds of potential advisors should realize that this approach
is unlikely to succeed.
(Last updated
.)