Laboratory for Intelligent Signal Processing
and Communications (LISPC)
Laboratory for Machine Intelligence and Neural and Soft Computing
(LMINSC)
Director
Rui J.P. de Figueiredo
Research Professor of Electrical Engineering and Computer Science,
and Mathematics
Located on the
campus of the University of California, Irvine,
the Laboratory operates under the auspices of the California Institute for
Telecommunications and Information Technology, in 4430 Calit2 Building. For
additional information, please contact
Professor Rui J.P. de Figueiredo, Office: 4417 Calit2 Building, University of
California Irvine, California 92697-2800, e-mail: rui at uci dot edu Tels:
Office: (949) 824-9953, Lab: (949) 824-7043, Fax: (949) 854-6528.
Some of the most important emerging signal, image, and information processing (SIIP) systems may
be viewed as distributed decision systems that require multi-scale nonlinear
dynamical system models to describe their complex behavior. SIIP systems are called intelligent and
cognitively agile when the underlying computational models incorporate functionality
commonly associated with computational intelligence and cognitive computing.
Under
the direction of Prof. de Figueiredo, the Laboratory is continuing pioneering
research (see Research Highlights, Selected
Publications I and Selected Publications II) on the mathematical
foundations as well as on modeling, algorithms, and architectures for Nonlinear (i.e., Not necessarily linear) Signal, Image and Information Processing (NSIIP) in
the context of emerging applications. These occur at two levels.
1. NSIIP: Filters and Processors
One component of the fundamental research is concerned
with mathematical definitions and properties of nonlinear recurrence and
nonlinear convolution in Hilbert Spaces of nonlinear functionals. In such a
formulation, the selection and appropriate use of the reproducing kernel of
such spaces plays a key role. The research is developing a novel methodology
for the modeling and design of generic nonlinear IIR (Infinite Impulse
Response) and FIR (Finite Impulse Response) models and filters for nonlinear
signal, image, and information processing (NSIIP), which can be customized for
applications.
Applications include (a) Pre-distorters for mitigation (by conventional and intelligent computing
approach) of nonlinear distortion
caused by high PAPR (Peak-to-Average-Power Ratio) in multi-carrier wireless
communication systems[1]
[75, 76], and (b) new nonlinear filters that maximize contrast on an image by taking human visual system
properties (expressed by Munsell’s
scale) into account [37]. Several other applications are described under Research Highlights and Selected
Publications I, and the results for others are in the process of being
submitted for publication.
2. NSIIP: Networks and Systems
The other component of the fundamental research is
concerned with networks, the nodes of which are NSIIP processors or
sub-networks of such processors, that is,
networks of networks. We
have created Hilbert Spaces where such nonlinear networks can reside, and thus be
optimally identified or designed by appropriate orthogonal projections of the
unknown desired network into the subspace spanned by the representers of the functionals
corresponding to input-output observations. The recent characterization of
fuzzy sets by Prof. de Figueiredo [38] [39] can help enhance this endeavor.
The above mathematical foundations and developments
enable us to model analyze and identify or design computationally intelligent and cognitively agile systems, that is, systems that are capable of adaptation, learning, with or
without supervision, evolution, discovery, and invention, often in an internet
environment. To execute such functionality, as allowed by our formulation, the
network structure usually needs to be complex; that is, multi-scale, parallel,
distributed, nonlinear, time-varying, robust, that is, capable of graceful
degradation, and in the case of design, affordable. Also, in most instances the
criterion for optimization is the conditional risk or some functional (i.
e., value assignment) equivalent to this risk in an approximation-theoretic sense.
Applications:
(a)
Ad-Hoc Networks. Strategies and methodology for optimal power
management and optimal power control in wireless communication networks [77] [78].
Insertion of intelligence and cognitive agility into such networks and extension
of the technology to Intelligent Multi-Media Communication (IMMC) systems,
under planning.
(b)
Intelligent
Search Engines. Under planning.
(c)
Telemedicine. Automated diagnosis of (i) dementia [29], and (ii)
diabetic retinopathy from retina images taken by a mydriatric camera (which
does not require pupil dilation) [67].
(d)
Intelligent
Multi-media Communications (IMMC). IMMC
systems will enable a machine to communicate with a human or with another
machine in the same way you and I do. For this purpose, a powerful abstraction
is needed to model, design, and implement machines that are capable of
detecting, classifying, and interpreting complex events present in the
multi-media signal. With this motivation, we are developing a rigorous abstract
approach to IMMC in human/machine systems, whereby the machine is modeled as a
Multiple-Input/Multiple-Output (MIMO) Intelligent Multi-Media Signal Processor
(IMMSP). The input-output map f of such a MIMO IMMSP resides in an appropriate
Reproducing Kernel Hilbert Space F of nonlinear functionals on the multi-media
input space X. The realization of the optimal model is obtained by an
appropriate orthogonal projection in F, subject to design-specification and
exemplary input-output-data constraints. Such optimal realizations of MIMO
IMMSPs naturally appear in the form of fuzzy neural systems [38] [39], the
fuzziness of which is encapsulated in the reproducing kernel of the space F
that gave birth to them. Work in progress.
(e)
Intelligent Data
Mining. This technology deals with
the segmentation of a very large database based on the conditional risk. In
other words, the criterion for optimization is the so-called “Lift”, i. e., the
ratio of the posterior to prior probabilities for the segment being searched.
Work has been completed but technology is not in the public domain.
3. Recent Members of the Group: Dr. Lin Fang, Dr. Bradley Denney, Dr. Katia Estabridis, Dr. Byung Moo Lee, Ms. Carolina Soto, Mr. Frederico Llarena
SOME
OTHER RECENT INFORMATION
- Dec. 10, 2007 - de Figueiredo elected to Russian Academy of Natural
Sciences, awarded honorary medal
Rui J. P. de Figueiredo, electrical engineering and computer science and mathematics research professor at UC Irvine, has been elected as a foreign member of the Russian Academy of Natural Sciences (RANS) U.S. Section and was awarded the academy's George V. Chilingar Medal of Honor for his important contributions to science and engineering. A UCI faculty member since 1990, de Figueiredo has been instrumental in strengthening the presence of the Institute of Electrical and Electronics Engineering's (IEEE) Circuits and Systems Society in Russia and Eastern Europe. - Professor Rui de Figueiredo is serving as Chair of the IASTED Ninth International Conference on Signal and Image Processing (SIP 2007), to be held in Honolulu, Hawaii on August 20 - 22, 2007. For details, see http://www.iasted.org/conferences/home-576.html
- Henry Samueli School of Engineering's Professor of Electrical Engineering and Computer Science, Rui J. P. de Figueiredo, will Co-Chair the IEEE 7th. Emerging Technologies Workshop: " Circuits and Systems for 4G Mobile Wireless Communications" to be held in St. Petersburg, Russia, on June 23-24, 2005. Prof. Alexander S. Korotkov of St. Petersburg State Polytechnical University will serve as the other Co-Chair, and Prof. H. Vincent Poor of Princeton University as the Advisory Board Chair.
- Prof.Rui J. P. de Figueiredo has been formally inducted as a Member of the International Informatization Academy (United Nations affiliated) and as an Honorary Member of the Russian Popov Society. Professor V. Shakhgildyan, Rector/President of the Moscow Technical University of Communications and Informatics, presented this prestigious commendation to Dr. de Figueiredo at the 2nd IEEE International Conference on Circuits and Systems for Communications, which was held in Moscow, Russia on June 30, 2004. See: http://www.eng.uci.edu class=MsoHyperlink>Scenes at the 2nd. IEEE International Conference on Circuits and Systems for Communications held in Moscow, Russia on June 30 - July 2, 2004. See http://www.mtuci.ru/info6/photo.htm
- Professor Rui J. P. de Figueiredo with Prof. Vagan Shakhgildyan (to the right), Rector of the Moscow Technical University of Communications and Informatics, and some other Conference Leaders, at the 2nd. IEEE International Conference on Circuits and Systems for Communications (ICCSC-2004) held in Moscow, Russia on June 30 - July 2, 2004.
- Professor R.J.P. de Figueiredo is to Co-Chair and give a plenary address at 2nd IEEE International Conference on Circuits and Systems for Communications (ICCSC-2004), to be held in Moscow, Russia on June 30 - July 2, 2004. More information at: http://www.mtuci.ru/info9.htm http://mw003.aast.edu/
- Professor de Figueiredo gave a plenary address at the 2003 IEEE International Midwest Symposium on Circuits and Systems held in Cairo, Egypt on December 28-31, 2003. The title of his lecture was "Nonlinear Signal Processing for Applications in Broadband Communications"
- Engineering Professor Wins Best Paper Award
- Professor Rui de Figueiredo's IEEE Distinguished Lecture at the Institute of Radio Engineering and Electronics, Russian Academy of Sciences, in July, 2000