[Home] [Back Issues] [Search] [Calendar] [Vol 1] [Vol 2] [Vol 3] [Vol 4] [Vol 5] [Vol 6]

The Wavelet Digest, Volume 6, Issue 1 (January 16, 1997)




Today's Editor: Wim Sweldens
                Lucent Technologies, Bell Laboratories
                wim@bell-labs.com



Today's Topics:

     1: Preprint: Thresholding Multiwavelets
     2: Preprint: Image compression via a statistical model of wavelets
     3: Preprint: Three preprints about Wavelet Noise Reduction 
     4: Preprint: Solving Hyperbolic PDEs Using Interpolating Wavelets
     5: Preprint: Plotting & Scheming with Wavelets
     6: Preprint: Wavelet Approximations on Closed Surfaces
     7: Preprint: Correlational Opponent Processing in Psychology
     8: Thesis:   A Statistical wavelet approach to model selection
     9: Meeting:  Multimedia Signal Processing Workshop
    10: Meeting:  Subband and Wavelet Transforms in Communications
    11: Course:   An Introduction to Wavelets
    12: Job:      Research Positions on wavelet image processing
    13: Job:      Professorship in Signal/Image Analysis
    14: CFP:      Special issue of JMP on Wavelets
    15: Contents: Const. Approx., Vol. 13, No. 1, 1997
    16: Answer:   Wavelet smoothing (WD 5.10 #22)
    17: Question: Wavelet-Pade
    18: Question: Time varying wavelet in image compression
    19: Question: Simulation of Speech Analysis bases on Wavelest
    20: Question: Optical Wavelet
    21: Question: Fractal image processing and scattering
    22: Question: Remote sensing image processing using wavelets
    23: Question: Biorthogonal wavelet transforms
    24: Question: Wavelet sampling 
    25: Question: Implementing Wickerhauser algorithms
    26: Question: Software for contrast enhancement
    27: Varia:    Change of publisher for Newland
     
Current number of members: 8453

E-mail:
   Help:                        help@wavelet.org
   Add yourself as a member:    add@wavelet.org
   Remove yourself as a member: remove@wavelet.org
   Publish in the next WD:      publish@wavelet.org

Visit us on the Web at http://www.wavelet.org/

Calendar of events:

 Jan 23-25: Workshop on Wavelets and Applications, Madras, India
 Mar 10-12: Workshop on Scientific Computing 97, Hong Kong
 Mar 20-22: Multiwavelets conference, Huntsville, TX
 Mar 21:    Wavelets & Subband transf. in Communications, NJIT
 Apr 13-16: 1997 ImageTech Conference Multimedia Imaging, Atlanta
*Apr 20   : Wavelet Transforms in Communications, Munich Germany
 Apr 22-24: SPIE Fourth Wavelet Application Conference, Orlando
 Jun 16-18: Engineering Applications of Neural Networks, Stockholm
 Jun 16-19: Sampling Theory and Applications, Aveiro, Portugal
*Jun 23-25: Multimedia Signal Processing Workshop, Princeton, NJ
 Jun 24-27: European conf. signal analysis and prediction, Prague
 Jun 25-27: Fractals in engineering, Arcachon, France
 Jul 2-4:   Scale-Space Theory in Computer Vision, Utrecht, Nl
 Jul 2-4:   Tutorial on Scale-Space Theory, The Netherlands
 Jul 25-Aug 1: SPIE Signal and Image Processing V, San Diego
 Aug 27-29: Time-Frequency and Time-Scale, Coventry, UK
 Sep 14-17: Time-Frequency Wavelet analysis symposium, Sacramento
 Sep 24-26: IEEE Neural Networks for Signal Processing, Florida
 Nov 16-Dec 6: UCLA short courses on communications, LA
 Dec 17-19: Introduction to wavelets, Boston







Date: Thu, 12 Dec 1996 11:51:31 +0000
From: Tim Downie <tim.downie@bristol.ac.uk>
Subject: #1 Preprint: Thresholding Multiwavelets

Title: The Discrete Multiple Wavelet Transform
and Thresholding Methods 

Authors: T.R.Downie & B.W.Silverman

Abstract: Orthogonal wavelet bases have recently been developed using
multiple mother wavelet functions.  Applying the discrete multiple
wavelet transform requires the input data to be preprocessed to obtain
a more economical decomposition.  We discuss the properties of
several preprocessing methods and their effect on the thresholding
results.  We propose a multivariate thresholding method for
multiwavelets where each wavelet coefficient vector is thresholded as
opposed to each element. A multiwavelet universal threshold is
obtained using the asymptotic behaviour of the $\chi ^2$
distribution. A simulation study demonstrates that when using the
Geronimo multiwavelet with an appropriate preprocessing method and
bivariate thresholding, better results are obtained compared with
using Daubechies wavelets.

A copy can be downloaded from 
http://www.stats.bris.ac.uk/~trd/threshmwave2.ps
or by emailing me at the address below

Thanks
Tim

  Tim Downie                          Department of Mathematics   
  Email: Tim.Downie@bristol.ac.uk     University of Bristol       



Date: Thu, 19 Dec 1996 20:37:16 -0500
From: Eero Simoncelli <simoncel@math1.cims.nyu.edu>
Subject: #2 Preprint: Image compression via a statistical model of wavelets

Dear Wavelet Digest readers,

I've just made available a preprint of our paper describing the joint
statistics of wavelet coefficients of natural images at adjacent
scales.  This is an extension of the work in my previously posted
preprint (WD, 7/12/96, vol5, num5: "Noise Removal via Bayesian Wavelet
Coring").

TITLE:  Progressive Wavelet Image Coding Based on a Conditional
	Probability Model.

AUTHORS: Robert W. Buccigrossi and Eero P. Simoncelli

URL:	http://www.cis.upenn.edu/~eero/publications.html

ABSTRACT: We present a wavelet image coder based on an explicit model
of the conditional statistical relationships between coefficients in
different subbands.  In particular, we construct a parameterized model
for the conditional probability of a coefficient given coefficients at
a coarser scale.  Subband coefficients are encoded one bitplane at a
time using a non-adaptive arithmetic encoder.  The overall ordering of
bitplanes is determined by the ratio of their encoded variance to
compressed size.  We show rate-distortion comparisons of the coder to
first and second-order theoretical entropy bounds and the EZW coder
(Shapiro, 1993).  The coder is inherently embedded, and should prove
useful in applications requiring progressive transmission.

Comments/criticism welcomed!

Eero Simoncelli



Date: Sat, 28 Dec 1996 11:16:56 +0100
From: Maarten Jansen <maarten.jansen@cs.kuleuven.ac.be>
Subject: #3 Preprint: Three preprints about Wavelet Noise Reduction 

The following three preprints are available from 
http://www.cs.kuleuven.ac.be/~maarten/Publications.html

1) TITLE: Generalized Cross Validation for wavelet thresholding

   AUTHORS: Maarten Jansen, Maurits Malfait and Adhemar Bultheel

   ABSTRACT:

Noisy data are often fitted using a smoothing parameter, controling
the importance of two objectives that are opposite to a certain
extent.  One of these two is smoothness and the other is closeness to
the input data.

The optimal value of this paramater minimizes the error of the result
(as compared to the unknown, exact data), usually expressed in the
$L_2$ norm.This optimum cannot be found exactly, simply because the
exact data are unknown.In spline theory, the Generalized Cross
Validation (GCV) technique has proven to be an effective (though
rather slow) statistical way for estimating this optimum.

On the other hand, wavelet theory is well suited for signal and image
processing. This paper investigates the possibility of using GCV in a
noise reduction algorithm, based on wavelet-thresholding, where the
threshold can be seen as a kind of smoothing parameter.  The GCV
method thus allows choosing the (nearly) optimal threshold, without
knowing the noise variance.

Both an original theoretical argument and practical experiments
are used to show this successful combination.

   NOTE: To appear in Signal Processing, 56(1)

   AVAILABLE AT:

http://www.cs.kuleuven.ac.be/~maarten/gcv.ps.gz

2) TITLE: Bayesian Approach To Wavelet-based Image Processing

   AUTHORS: Maurits Malfait, Maarten Jansen and Dirk Roose

   ABSTRACT:

We present a new method for the reduction of noise in images, using a
wavelet transform. The method relies on two principles. The first is
the characterization of the local function regularity by wavelet
coefficients.The second is an a priori, geometrical model for wavelet
coefficients.  Both are combined in a Bayesian framework, to compute
for each wavelet coefficient the probability of being ``sufficiently
clean''.  The manipulation of the wavelet coefficients is consequently
based on the obtained probabilities.

   NOTE: presented at the Joint Statistical Meetings, Chicago,
August             1996

   AVAILABLE AT:

http://www.cs.kuleuven.ac.be/~maarten/bayaug.ps.gz

3) TITLE: Multiple wavelet threshold estimation by generalized
cross               validation for data with correlated

   AUTHORS: Maarten Jansen and Adhemar Bultheel

   ABSTRACT:

De-noising algorithms based on wavelet thresholding replace small
wavelet coefficients by zero and keep or shrink the coefficients with
absolute value above the threshold. The optimal threshold minimizes
the error of the result as compared to the unknown, exact data. To
estimate this optimal threshold, we use Generalized Cross
Validation. This procedure does not require an estimation for the
noise energy.  Originally, this method assumes uncorrelated noise. In
this paper we describe how we can extend it to images with correlated
noise.

   AVAILABLE AT:

http://www.cs.kuleuven.ac.be/~maarten/corgcv.ps.gz



Date: Mon, 13 Jan 1997 14:28:33 +0100 (MET)
From: Mats Holmstr|m <matsh@tdb.uu.se>
Subject: #4 Preprint: Solving Hyperbolic PDEs Using Interpolating Wavelets

Title : Solving Hyperbolic PDEs Using Interpolating Wavelets

Author: Mats Holmstr=F6m

Abstract.  A method is presented for adaptively solving hyperbolic
PDEs.  The method is based on an interpolating wavelet transform using
polynomial interpolation on dyadic grids.  The adaptability is
performed automatically by thresholding the wavelet coefficients.
Operations such as differentiation and multiplication are fast and
simple due to the one-to-one correspondence between point values and
wavelet coefficients in the interpolating basis.  Treatment of
boundary conditions is simplified in this sparse point representation.
Numerical examples are presented for one- and two-dimensional
problems.  It is found that the underlying finite difference method's
order of convergence is preserved, and that the proposed method
outperforms a finite difference method for certain problems in terms
of flops.

The full report is available at =

ftp://ftp.tdb.uu.se/pub/reports/report-189-1996.ps.gz

Sincerely,

Mats Holmstr=F6m                                   matsh@tdb.uu.se  =

Uppsala University                               http://www.tdb.uu.se/~ma=
tsh =

Dept. of Scientific Computing                    Ph.: +46 18 18 29 80  =

Box 120, S-751 04 Uppsala, Sweden.               Fax: +46 18 52 30 49  =



Date: Tue, 14 Jan 1997 10:39:52 -0500
From: colm@mathcs.emory.edu (Colm Mulcahy)
Subject: #5 Preprint: Plotting & Scheming with Wavelets

Title: Plotting & Scheming with Wavelets

This is an elementary, linear algebra-based intro to wavelets aimed at
a mathematical audience with little sophistication.  Only Haar
wavelets are considered in detail, with image compression as the
motivation.  MRA's are introduced as the way forward to Daubechies
wavelets etc.

It is published in the December 1996 issue of Mathematics Magazine, by
the Mathematical Association of America.

We were much inspired by Stollnitz, DeRose, and Salesin's fine
wavelets primers (much of the contents of which resurfaced in their
recent book).
 
Plotting & Scheming with Wavelets can be found in postscript form at
http://www.spelman.edu/~colm/papers.html, as can associated Matlab
m-files, along with an even simpler introduction to wavelet
compression which any 15 year old could understand.



Date: Thu, 16 Jan 1997 07:39:06 -0500
From: Frank Schneider <schneider@mathematik.uni-kl.de>
Subject: #6 Preprint: Wavelet Approximations on Closed Surfaces

Authors: W. Freeden, F. Schneider

Title:   Wavelet Approximations on Closed Surfaces
         and Their Application to Boundary-value Problems
         of Potential Theory

Abstract: Wavelets on closed surfaces in Euclidean space $ \R^3 $ are
introduced starting from a scale discrete wavelet transform for
potentials harmonic down to a spherical boundary. Essential tools for
approximation are integration formulas relating an integral over the
sphere to suitable linear combinations of functional values
(resp. normal derivatives) on the closed surface under consideration.
A scale discrete version of multiresolution is described for potential
functions harmonic outside the closed surface and regular at
infinity. Furthermore, an exact fully discrete wavelet approximation
is developed in case of band-limited wavelets.  Finally, the role of
wavelets is discussed in three problems, namely (i) the representation
of a function on a closed surface from discretely given data, (ii) the
(discrete) solution of the exterior Dirichlet problem, and (iii) the
(discrete) solution of the exterior Neumann problem.

This preprint is available at

http://www.mathematik.uni-kl.de/~wwwgeo/agtm171.ps

or

http://www.mathematik.uni-kl.de/~wwwtecm/preprints/reports/report_171.ps.gz

Best regards

W. Freeden, F. Schneider

Prof. Dr. Willi Freeden               email: freeden@mathematik.uni-kl.de
Dipl.-Math. Frank Schneider           email: schneider@mathematik.uni-kl.de  
Department of Mathematics             fax:  ++49 631 29081
Geomathematics Group                  http://www.mathematik.uni-kl.de/~wwwgeo
University of Kaiserslautern
P.O. Box 3049
D-67653 Kaiserslautern
Germany



Date: Wed, 11 Dec 1996 11:12:28 -0500 (EST)
From: Ron Blue <rcb1@lex.lccc.edu>
Subject: #7 Preprint: Correlational Opponent Processing in Psychology

robot at http://www.aston.ac.uk/~batong/Neutronics
which uses CORE (TM) Processing or Correlational Opponent Ratio
Enhanced Processing or wavelets to learn.

Unification model:  Correlational Opponent Processing
Available by email request only.

KEYWORDS: correlational opponent-processing, opponent process,
wavelets, neuro net, excitatory, inhibitory, EXIN, oscillation,
holographic, eigenfunction, chaos, nerves, habituation, discorrelation,
memory, sensations, perceptions, emotions, evolution, brain damage

Abstract:

The correlational opponent-processing theory using wavelets, quasi-
holographic memory and eigenfunction equivalence generates new insights
into many areas of psychology.  The model seems especially strong in
joining many contradictory scientific facts into an unifying whole.
Significant implications from the model exist for a wide range of
psychological topics and principles.

The correlational opponent-processing theory is a neuro homeostasis
integration psychological immune theory that would connect phenomena
such as sensation, perception, movement, habituation, memory,
representations, learning, cognition, personality, psychopathology,
paradoxical integration, emotion, and evolution of the mind under a
unified theory.

All brain activity may be viewed as an effort to assimilate and
accommodate all experience into neuro-energy-efficient eigenfunction
equivalence or quasi-holographic correlational opponent-processing
recordings.

Stimuli causes brain wave modulations which interact with carrier or
reference wavelets.  This interaction creates a quasi-holographic
stimulus wavelet.  The opponent-process creates an opposing quasi-
holographic memory wavelet.  Through this process the correlations or
associations of experience are encoded to memory.  Every wavelet,
regardless of source or type, triggers an opposing wavelet.  The
function of the opposing wavelet or feedback is to diminish the
intensity of neural processing.  A wavelet potential is stored or hard
wired as long-term potentiation opponent-processes in nerve cells and
the interconnections between nerve cells.  The wavelets are quasi-
holographic and allow recovery of information due to the interaction
of reference carrier wavelets and stimuli, thought, motor movement,
and emotional arousal.



Date: Tue, 24 Dec 1996 11:50:16 -0600 (CST)
From: Geung-Hee Lee <ghlee@picard.tamu.edu>
Subject: #8 Thesis: A Statistical wavelet approach to model selection

My Ph.D. thesis, entitled:

"A Statistical wavelet approach to model selection and data driven
 Neyman smooth tests."

It can be obtained by anonymous ftp to

  ftp://stat.tamu.edu/pub/aliaw/disset.ps.gz.

ABSTRACT

Wavelets have received a lot of attention in statistics 
since Donoho and Johnstone(1994, Biometrika) introduced wavelet shrinkage 
estimators, which include two important ideas -- wavelets as a new local 
basis and thresholding. From these ideas, we can reilluminate the previous
 methods based on Fourier series and truncated estimators.
In this dissertation, we develop model selection criteria in 
the orthogonal series estimator and
data driven Neyman smooth tests as lack-of-fit tests using these two ideas.

Regression analysis can be explained as an iterative cycling procedure
between model selection and lack-of-fit testing of the selected model.
We develop new model selection criteria to detect low as well as
dominant high frequency behavior for a truncated series estimator. For
the wavelet series estimator, we propose a rule to choose a threshold
and a truncation point simultaneously. This estimator is shown to
provide better performance in terms of visual quality and mean square
error compared with wavelet series estimators with fixed truncation
points such as $RiskShrink$ and $VisuShrink$.

Data driven Neyman smooth tests are used as omnibus tests to detect
misspecification error of the model.  The performance of data driven
Neyman smooth tests depends on the selected smoothing parameter, basis
and the form of orthogonal series estimator.  In this dissertation, we
modify the data driven Neyman smooth test in order to improve its
performance for a wider class of alternatives using different model
selection criteria, different bases such as wavelets, and different
estimators.

GeungHee Lee.
Statistics, Texas A&M



Date: Fri, 06 Dec 1996 16:41:19 -0500
From: Tsuhan Chen <tsuhan@research.att.com>
Subject: #9 Meeting: Multimedia Signal Processing Workshop

IEEE Signal Processing Society
1997 Workshop on Multimedia Signal Processing
June 23-25, 1997
Princeton, New Jersey, USA

SECOND CALL FOR PAPERS

Thanks to the sponsorship of IEEE Signal Processing Society, the First
IEEE Workshop on Multimedia Signal Processing (MMSP) will be held at
Princeton University, Princeton, New Jersey, USA, on June 23-25, 1997.
MMSP strives to bring together people in the signal processing
society, the communications society, and the computer society, and to
offer a forum for interaction.  This workshop will feature keynote
addresses, technical presentations, panel discussions, exhibit tours,
and special sessions by invited speakers.  To encourage interaction,
we are planning sessions with the format of presentation followed by
technical discussion on the same topic.  These topics include, from a
signal-processing viewpoint: integration of media, human/machine
interfaces, and multimedia networks.  In particular, papers are
solicited for, but not limited to, the following general areas:

1 Multimedia Processing: Compression for Multimedia (Audio, Video,
Speech, etc.), Integration of Media, Joint Audio-Video Processing

2 Multimedia Databases: Indexing, Retrieval, and Archiving, Authoring
and Editing, Digital Library, Content Preparation and Presentation

3 Multimedia System Design and Implementation: Parallel Architecture,
ASIC Design, Software and Hardware Design, System Integration, Signal
Acquisition

4 Human-Machine Interface and Perception: Content
Recognition/Analysis/Synthesis, Speech Recognition and Synthesis,
Audio/Music Signal Processing and Synthesis, Multimodal Interaction
(Speech, Image, and Video), Audiovisual Perception Quality and Human
Factors, Cognitive Sciences

5 Multimedia Communications: Equalization and Synchronization,
Transport Protocols, QoS Control, Error Concealment and Loss Recovery,
Rate Control and Hierarchical Coding, Wireless Communication

6 Multimedia Applications: WWW and Hypermedia, Videoconferencing and
Collaboration Environment, Education and Distant Learning,
Telemedicine, Home-Shopping, Gaming and Virtual Reality, SDTV, HDTV,
SHDTV and Video on Demand

7 Standards and Related Issues: ITU-T Standards for Audiovisual
Communication, MPEG-1, MPEG-2, MPEG-4, MHEG, DAVIC, ATM Forum, IETF,
HTML, VRML, and others, Compressed Domain Processing

SUBMISSION PROCEDURES

Prospective authors are invited to submit 5 copies of extended
summaries of no more than 4 pages. The top of the first page of the
summary should include a title, authors' names, affiliations, address,
telephone and fax numbers and e-mail address if any.  Camera-ready full
papers (each up to 4 pages) of accepted proposals will be published in
a proceedings and distributed at the workshop.  For further
information, please contact Ray Liu or visit our home page at
http://neuro.ece.wisc.edu/~tsuhan/mmtc.html  Please send paper
submissions to:

Amy Reibman, Room 4E520
AT&T Labs - Research, 101 Crawfords Corner Rd, Holmdel, NJ 07733-3030,
USA, Phone +1-908-949-3470, Fax +1-908-949-3697

Text-only abstracts may be submitted electronically to:
amy@research.att.com

SCHEDULE

Extended summary received by:                          December 20, 1996
Notification of acceptance:                            February 1, 1997
Camera-ready accepted papers received by:              March 15, 1997
Advanced registration received before:                 May 15, 1997

DEMONSTRACTION

We also call for proposals for demonstration, exhibits, and company
tours.  
Please contact M. Reha Civanlar at civanlar@research.att.com

Co-Chairs

Tsuhan Chen AT&T Labs - Research tsuhan@research.att.com
Sun-Yuan Kung Princeton University kung@ee.princeton.edu

Technical Co-Chairs

Amy Reibman AT&T Labs - Research amy@research.att.com
Fred Juang Lucent Technologies bjuang@lucent.com

Local Chair

Mike Orchard Princeton University orchard@ee.princeton.edu

Publication Chair

Yao Wang Polytechnic University yao@vision.poly.edu

Finance Chair

Shih-Fu Chang Columbia University sfchang@ctr.columbia.edu

Exhibit Chair

M. Reha Civanlar AT&T Labs - Research civanlar@research.att.com

Publicity Chair

Ray K.J. Liu Univ. of Maryland, College Park kjrliu@isr.umd.edu

Registration Chair

Yu Hen Hu University of Wisconsin, Madison hu@eceserv0.ece.wisc.edu

Program Committee

A. N. Akansu, NJIT 
E. B. George, TI
A. Hsu, Siemens
J.-N. Hwang, UW at Seattle
C.-C. J. Kuo, USC
S.-P. Liou, Siemens
A. Katsaggelos, Northwestern U.
C.-H. Lee, Lucent Technology
T. Meng, Stanford University
R. Nakatsu, ATR
T. Nishitani, NEC
P. Pirsch, Univ. of Hannover
D. Shaver, TI
B. Sheu, USC
M. J. T. Smith, Georgia Inst. Tech.
J. Sorensen, Tech. U. of Denmark
G. Sullivan, PictureTel
M.-T. Sun, UW at Seattle
A. M. Tekalp, U. of Rochester
P. Westerink, IBM
A. Zakhor, UC Berkeley
Y.-Q. Zhang, David Sarnoff

Organized by the Multimedia Signal Processing Technical Committee, in
   cooperation with:
   Digital Signal Processing Technical Committee
   Image and Multidimensional Signal Processing Technical Committee
   Speech Processing Technical Committee, and
   New Jersey Center for Multimedia Research



Date: Tue, 10 Dec 1996 14:32:37 -0500
From: ali akansu fac ee <ali@hertz.njit.edu>
Subject: #10 Meeting: Subband and Wavelet Transforms in Communications

     IEEE 1997 International Conference on Acoustics, Speech,
        and Signal Processing (ICASSP'97) Munich, Germany

                      A Half-day Tutorial
                              on 

          SUBBAND & WAVELET TRANSFORMS IN COMMUNICATIONS

  Ali N. Akansu, NJIT (USA) & Mark J.T. Smith, GeorgiaTech (USA)

             April 20, 1997 (Sunday), 9:00-12:00 AM
             Forum Hotel (F/London), Munich

 This tutorial aims to focus on the recent developments of subband and
 wavelet transforms for digital communications. The fundamentals,
 linkages and distinctions of subband and wavelet transforms will be
 presented.  The design methodologies for subband and wavelet
 transform bases will be introduced. Then, the state-of-the-art
 applications of subband transforms and their extensions in
 interference excision, spread spectrum communications, discrete
 multitone (DMT) modulation, code division multiple access (CDMA) and
 low probability of intercept (LPI) communications along with subband
 (wavelet) image/video compression techniques will be presented. The
 applications specific requirements and their connections with the
 fundamentals of subband and wavelet transform theories will be
 highlighted.

 First part of the tutorial will visit the fundamentals of wavelet and
 subband transform theories. This will help the audience to appreciate
 the linkages as well as the distinctions of these popular linear
 transform techniques. In the second part, we will present the
 state-of-the-art digital communications applications of subband
 transforms. These recently emerging applications will be interpreted
 from subband transform theory point of view.  An evaluation and
 assessment for potential improvements of these digital communications
 techniques will be made.

 INTENDED AUDIENCE:

 Signal processing and communications engineers, applied
 mathematicians, graduate students and managers monitoring
 the new potential technologies in digital communications.

 OUTLINE:

I. Subband and Wavelet Transforms: Fundamentals, 
   Linkages and Distinctions
   A. Introduction and Overview
   B. Subband Transform Theory
   C. Wavelet Transform Theory
   D. Subband Transforms in Wavelet Theory
   E. Distinctions of Wavelet and Subband Transforms
   F. Dyadic Subband Tree as a Fast Wavelet Transform
   G. Design of Subband and Wavelet Transform Bases

II. Subband Transforms in Digital Communications 
   A. Time-Frequency Interpretation of Subband Bases
   B. Subband Trees (Wavelet Packets): A Flexible Linear Transform Tool
   C. Smart Time-Frequency Exciser in Spread Spectrum Communications
   D. Discrete Multitone (DMT) Modulations and Optimal Basis Design
   E. A Novel Code Division Multiple Access (CDMA) Communications
      System as an Extension of Subband Theory
   F. LPI Communications Using Subband Transforms
   G. Subband (wavelet) Transforms in Multimedia Communication
   H. Discussions and Future Directions 



Date: Tue, 14 Jan 1997 11:00:46 -0500
From: colm@mathcs.emory.edu (Colm Mulcahy)
Subject: #11 Course: An Introduction to Wavelets

"An Introduction to Wavelets", a short course, will be given by Colm
Mulcahy from 2:00 to 5:00 PM on Thursday 13th March, at Georgia Tech,
during the 75th Anniversary MAA (Math Assoc of America) Southeastern
Section Meeting.

This meeting will be held jointly at Spelman College and Georgia Tech,
Atlanta, Georgia, from 13-15 March 1997.

$20 + registration form must reach the MAA secretary-treasurer by 24
February: Stephen Davis, CPO#274, Davidson College, Davidson, NC
28036-1719.

($20 must be in the form of a check made mayable to the MAA;
registration form is available from:
http://www.huntingdon.edu/MAA/97_Announcement.html)

(if course gets enought pre-registrated folks to run, and spaces
remain, you can still register at the meeting itself)

Colm Mulcahy
Dept. of Mathematics
http://www.spelman.edu/~colm/



Date: Mon, 9 Dec 1996 11:36:07 -0500
From: sheng@phy.ulaval.ca (Yunlong Sheng)
Subject: #12 Job: Research Positions on wavelet image processing

    Research Positions on wavelet image processing in Canada

    Research Associate, Post-Doctoral fellow and graduate research assistant
positions on wavelet image processing are now available at Laval
University, Center for Optics Photonics and Laser, image science
group, and at the company Lockheed Martin Electronic System Canada
(LMESC) for a collaborative research project on multi-spectrum image
stabilization, registration and fusion using the wavelet
transform. The research will involve wavelet image feature extraction,
motion estimation and multiresolution image fusion.
    The project is funded for three years by the Department of
National Defense (DND), the Natural Sciences and Engineering Council
of Canada (NSERC) and LMESC through a DND-NSERC-industry research
partnership program.
    Candidates should have a Ph.D. or a B.Sc. in Electrical Engineering,
Computer Science, Applied Mathematics or Physics, and strong skill for
programming, computer algorithm analyzing and problem
solving. Knowledge in image processing and wavelet transform will be
very much appreciated.
    One Post-Doctoral or Research Associate will work in the R&D
department at LMESC (located in Montreal).
    Salary is according to candidate skills and the NSERC rules.
    Hiring conditional to contractual agreements between all parties
(hopefully in January 1997).
    We invite interested candidates to contact:

Prof. Yunlong Sheng
Physics Department 
Laval University
Ste-Foy, Quebec, Canada G1K 7P4
Tel: 418-656-3908
Fax: 418-656-2623
E-mail: sheng@phy.ulaval.ca

    Research Associate and/or Post-Dotoral candidates can also send their
CV to:

Dr. Langis Gagnon
R&D Department
Lockheed Martin Electronic Systems Canada
6111 Ave. Royalmount
Montreal, Quebec, Canada H4P 1K6
Tel: 514-340-8310 ext. 8715
Fax: 514-340-8318
E-mail: langis.gagnon@lmco.com



Date: Thu, 2 Jan 1997 14:27:04 +0100 (MET)
From: Alfred Hanssen <Alfred.Hanssen@phys.uit.no>
Subject: #13 Job: Professorship in Signal/Image Analysis

		  UNIVERSITY OF TROMSO, NORWAY

	Institute for Mathematical and Physical Sciences
	   Section for Physics, Applied Physics Group

		  ASSOCIATE PROFESSORSHIP(S) IN

	    S i g n a l - / I m a g e   A n a l y s i s

				and

    E l e c t r o n i c s / M e a s u r e m e n t   T e c h n i q u e s

At the world's northernmost university (in northern Norway) we are
happy to announce up to two vacant positions in signal-/image analysis,
and up to two positions in electronics/measurement techniques.

Scientists with a background in one or more of the following fields:
Satellite remote sensing, one-/multidimensional signal processing, 
image analysis, statistical signal analysis, and/or time series analysis,
are invited to apply for the vacancies in signal-/image analysis.

Scientists with experimental background in one or more of the
following fields: Measurement techniques, physical electronics,
microwaves, sensor technology, and/or (electro)optics, are invited to
apply for the vacanies in electronics/measurement techniques.

The Applied Physics Group has the responsibility for an engineering
program in Signal Processing and Electronics.  The vacant positions
are connected to this engineering program, and the appointees are
expected to play an active role in research and teaching activities
connected to this program.  It will be possible for the succesful
candidate to initiate his/her own research activities.  A PhD degree
is required for the positions, and a postdoctoral record is considered
desirable.

The annual salary is at present NOK 280,531 - 302,031 (1 USD = 6.4
NOK).  A full sabbatical year is granted every fifth year.  The
employees are expected to use 50% of the time to their own research,
and 50% of the time to teaching and administration.

For further information about the positions, and for information about
the formalities involved in the search process, please contact
Associate Professor T. Eltoft (pcte@phys.uit.no) (signal processing)
or Professor A. van Rheenen (arthur@phys.uit.no) (electronics).  The
deadline for applications is January 31, 1997.



Date: Wed, 18 Dec 96 11:21:05 +0100
From: Bruno Torresani <torresan@cptsx1.univ-mrs.fr>
Subject: #14 CFP: Special issue of JMP on Wavelets

		     ANNOUNCEMENT

		  JMP Special Issue

       	  WAVELETS and TIME-FREQUENCY ANALYSIS

In April, 1998, the Journal of Mathematical Physics will publish
a special issue on  Wavelet and Time-Frequency  Analysis.

Wavelet analysis and, more generally, time-frequency analysis, have
emerged in the last ten years as a true interdisciplinary
subject. Among the main contributors were mathematical physicists, as
well as pure mathematicians, signal and image processors and
electrical engineers.  Various approaches to time-frequency analysis
have been developed by different schools, emphasizing different
aspects relevant for specific applications.

One of the goals of the special issue is to review the connections
between wavelets and mathematical physics, from the point of view of
the foundations of the theory as well as for applications.  Several
facets and applications of wavelet analysis will be discussed in this
volume, ranging from algebraic and geometrical ones (for example, the
group theoretical aspects, and the connection to coherent states and
quantization) to analytic aspects (for example, mathematical analysis
of partial differential equations, or fractal and multifractal
analysis).  Some related topics will also be addressed: for example,
wave packets and the Schrodinger equation, wave propagation, Wigner
functions, non-stationary stochastic processes, etc.

The special issue will contain a few review papers devoted to the main
approaches to and applications of wavelets, as well as original
research papers. Most contributions will be invited papers, but
submissions are welcome as well. All contributions will be reviewed by
anonymous referees. The final authority for the acceptance of papers
is the Edeitor of JMP, Roger Newton. The special Editor of this issue
will be

Dr Bruno Torr\'esani
CPT, CNRS-Luminy, Case 907
13288 Marseille Cedex 09
France
Phone: (33) 04 91 26 95 23, Fax: (33) 04 91 26 95 53
e-mail: torresan@cpt.univ-mrs.fr

Manuscripts and inquiries should be sent directly to him. It will
facilitate handling if authors send manuscripts to the special Editor by
electronic mail, in addition to three hard copies. To meet the
April publication date, the papers should be in the hands of the special
Editor no later than August 15, 1997.



Date: Tue, 31 Dec 1996 20:36:50 -0500 (EST)
From: "Prof. E.B. Saff" <esaff@math.usf.edu>
Subject: #15 Contents: Const. Approx., Vol. 13, No. 1, 1997

1	P. Wojtaszczyk
	On Unconditional Polynomial Bases in Lp and 
	Bergman Spaces

17	K. Yu. Osipenko
	Exact n-Widths of Hardy-Sobolev Classes

29	S. Dahlke, V. Latour, and M. Neeb
	Generalized Cardinal B-Splines: Stability, Linear
	Independence, and Appropriate Scaling Matrices

57	G. Davis, S. Mallat, and M. Avellaneda
	Adaptive Greedy Approximations

99	Z. Ditzian and D.S. Lubinsky
	Jackson and Smoothness Theorems for Freud Weights
	in Lp (0 < p \le \infty)

RESEARCH PROBLEMS

153	D. Gaier
	On the Convergence of the Bieberbach Polynomials
	Inside the Domain: Research Problems 97-1



Date: Mon, 9 Dec 1996 09:41:56 +0100 (MET)
From: Francesco P. Lovergine <loverg@ercole.iesi.ba.cnr.it>
Subject: #16 Answer: Wavelet smoothing (WD 5.10 #22)

> From: Christian Oehreneder <coehrene@geoinfo.tuwien.ac.at>
> Subject: #22 Question: I) smoothing on the interval / II) an integral
> 
> Dear wavelet-colleagues!
> 
> I need help on tow wavelet-related problems.
> 
> I) I need an efficient smoothing algorithm for functions defined on an
> interval in R or R^2. One but not the only way would be a wavelet
> basis for those domains.  So far I have used the Lifting-Scheme, but I
> want to test others too.  What kind of solutions do exist on this
> problem?  Has anybody got computer-code, so that I can try it on my
> problem?
> 

Wavelet smoothing is not the only scheme of smoothing, of course.  A
different class of smoother are based on regularization theory a la
Thikonov. See Tomaso Poggio and many others for applications in image
processing and computer vision. These method are roughly based on the
definition of an energy functional to minimize, which measures both
data compatibility of the final solution and its smoothing.  The
resulting problem is well-posed and is expressed in terms of a PDE
problem.  Classical models used are the membrane and thin-plate model.
More complex scheme (with non-convex functionals) can be defined to
integrate discontinuities in the solution. This scheme are not
well-posed, generally.  For numerical methods see D.Terzopoulos and
many others.

Some references:

author = "T. Poggio and V. Torre and C. Koch",
title = "Computational Vision and Regularization Theory",
journal = "Nature",
volume = 317,
year = 1985,
pages = "314--319"

@article{bib:Terzopoulos86b,
author="D. Terzopoulos",
title="Regularization of Inverse Visual Problems Involving
Discontinuities",
journal=PAMI, year=1986, volume=8, number=4, month=jul, pages={413--424}
}

@article{bib:Terzopoulos88,
author="D. Terzopoulos",
title="The Computation of Visible Surface Representation",
journal=PAMI, volume=10, number=4, month=jul, pages={417--438},
year=1988
}

@book{bib:Tikhonov77,
author="A.N. Tikhonov and V.Y. Arsenin",
title="Solutions of Ill-posed Problems",
publisher="W.H. Winston ed.", year=1977
}

    f.lovergine@iesi.ba.cnr.it 
    CNR-IESI               Fax: +39 80 5484311
    Via Amendola 166/5     Pho: +39 80 5481969
    I-70125 BARI           I T A L Y



Date: Tue,  7 Jan 1997 22:01:23 +0300 (UKR)
From: "Victor Mikolenko" <root@vit.alkar.dp.ua>
Subject: #17 Question: Wavelet-Pade

Wavelet-Pade transform, comparison of Fourier-Pade transform and wavelet

Dear colleagues,

I am layman in the field of wavelet transform, but often use Fourier
transform.  My first question is: For Fourier transform we may use
Pade approximants and improve it. Is it possible to do something
similar for wavelet transform?  Second question is: As I understand,
non-smoothing is the most problem for wavelet transform (is it rihgt,
by the way?). On the other hand, Fourier transform gives us smooth
solution, but the rate of convergence is small. May be, if we use
Pade-Fourier transform, we may overcome this difficulty and obtain
smooth transform with rate of convergence similar to the wavelet?
Sincerely yours I.Andrianov

E-mail: root@vit.alkar.dp.ua



Date: Tue, 10 Dec 1996 11:20:44 +0100 (MET)
From: nacer.bounoua@univ-valenciennes.fr (nbounoua)
Subject: #18 Question: Time varying wavelet in image compression

Dear wavelet experts,

I'm in thesis of PhD, doing research work on image compression with
time varying filter banks and wavelets. I'd like to know where I can
get the papers on this subject . The more, the better.

Best regards

nacer bounoua
Dpt. O.A.E.
Universit=E9 de Valenciennes
B.P.311 - 59304 Valenciennes Cedex
France
phone: (3) 27 14 14 50
fax: (3) 27 14 11 89



Date: Wed, 11 Dec 1996 18:19:16 -0800 (GMT)
From: Renganathan a/l Palanisamy <b36580@stud.upm.edu.my>
Subject: #19 Question: Simulation of Speech Analysis bases on Wavelest

Dear Waveleters,

I am currently involved in simulating a model of speech analysis
system based on Pitch Synchronous wavelet representation of speech
signals.  I would appreciate it if anyone can assist me in
identifying the important aspects,giving pitch estimation algorithms
and softwares(affordable) which I can use to simulate the proposed
system also in guiding me as to how to obtain synthetic speech files
to test the system.  At the moment,I have only Matlab ver. 4.0 without
any toolbox.

My dateline is 28 February 1996. So please respond as soon as possible.

Your time and effort is greatly appreciated.

Yours Sincerely,

Renganathan S/O Palanisamy
E-mail : b36580@stud.upm.edu.my
Course : B.Sc(Hons) Physics
Project : Final Year : Simulation Of Speech Analysis System
Univ   : Physics Dept
         Faculty Of Science & Environmental Studies
         Malaysian Agricultural University. 	    	 



Date: Fri, 13 Dec 1996 14:35:27 +0000
From: Radhakrishnan Periasamy <r.periasamy@ecs.soton.ac.uk>
Subject: #20 Question: Optical Wavelet

I will be happy if someone could help me in answering the following
questions. 
What is optical wavelet? Where it is used? and How it is different?

Thnaks
RK

rk94r@ecs.soton.ac.uk



Date: Fri, 20 Dec 1996 14:59:59 +0800
From: "optoinfo@blue.hust.edu.cn" <optoinfo@blue.hust.edu.cn>
Subject: #21 Question: Fractal image processing and scattering

I am presently engaged in a research problem that involves optical
imaging the objects hidden inside high scattering media.  I wonder if
the fractal image processing methods could be used to substract the
effectness of multi-scatterring between local inhomogenities.

Mikehust
optoinfo@blue.hust.edu.cn



Date: Fri, 3 Jan 1997 16:26:29 -0000
From: Peixin Hou <p.hou@ee.surrey.ac.uk>
Subject: #22 Question: Remote sensing image processing using wavelets

I am interested in remote sensing image compression, classification
and object recognition using wavelet or other multi-resolution
method. Are there some papers,research or programs on it?  Thanks.

Reply to: P.Hou@ee.surrey.ac.uk



Date: Thu, 9 Jan 1997 09:46:06 +0530
From: Santhosh <san@ee.iitm.ernet.in>
Subject: #23 Question: Biorthogonal wavelet transforms

Dear waveleteers,
      I am a postgraduate research scholar in the area of "Wavelet Transforms
for Low bit rate Image coding". I implemnted orthogonal wavelets and
did VQ to achieve good compression ratios for still images. But, as it
has been proved that biorthogonal wavelets are better for Image coding
appliccation becuase of linear phase, currently I am looking into
biorthognal bases. However, I am encountering lots of problems in
implementing these. Perhaps, my question is too trivial. Please
forgive me. But, if you have any software which implements the Forward
and Inverse wavelet transforms using biorthogonal bases, it will be of
immense use for me. Thanks in advance...

                                                      Yours sincerely,
                                                       N Santosh



Date: Thu, 9 Jan 1997 22:48:04 GMT
From: Julie Wilson <julie@yorvic.york.ac.uk>
Subject: #24 Question: Wavelet sampling 

I am a beginner to wavelets and would like to ask a basic question
about discrete orthogonal wavelets. As far as I can see the theory
(according to people like Mallat and Daubechies) seems to apply to 
the coefficients <f,phi> where f is the signal and
phi is the scaling function and rather than the signal
itself. However, other papers and, in particular, the book
"Numerical recipes in Fortran" which gives code to do a wavelet transform
using Daubechies wavelets, seem to use the signal, f, rather than <f,phi>.
I have tried to work out what you are doing if you do this, thinking of
f as f(n) = sum_k <f,phi_k> phi(n-k) but it just doesn't seem to give 
anything sensible. Does anyone understand this question and if so can
they be bothered to answer it?

confused of York, a.k.a julie@yorvic.york.ac.uk

[Answer]
[Answer]



Date: Mon, 13 Jan 97 11:04:41 UT
From: "Paul Stringer" <idsldev@msn.com>
Subject: #25 Question: Implementing Wickerhauser algorithms

Question

Has anyone successfully implemented the algorithms in Wickerhauser's
"Adapted Wavelet Analysis From Theory To Software"? If so, I would
like to get in touch as I have a few questions to ask. I have got as
far as chapter 8, and I've been able to draw time-frequency plots, but
I have had a few problems and although my results look right I haven't
been able to test them thoroughly. At present I'm a bit stuck with how
to implement the best basis algorithm and how it links to the wavelet
packet approach detailed in chapter 7.

Has anyone tried to implement the algorithms and failed? Maybe we
could discuss. The limited success I have had has taken two months of
work so far.  I'm fairly close to being finished, so it would be a
shame to hit a dead-end now.

Yours,
Toby Sharp
idsldev@msn.com



Date: Mon, 13 Jan 1997 20:03:37 +0200 (EET)
From: grad0060@di.uoa.gr
Subject: #26 Question: Software for contrast enhancement

I am looking for source code, which implements contrast enhancement
(based on multiscale edges) on medical images with the help of
wavelets.

Thanks,

Alexis Savopoulos

Email: grad0060@di.uoa.gr



Date: Sun, 22 Dec 1996 19:37:47 GMT
From: David Newland <den@eng.cam.ac.uk>
Subject: #27 Varia: Change of publisher for Newland

Recently I was contacted about the apparent non-availability of my
books in the US. This is not the case, but the publisher and ISBN
numbers for the USA have changed.  The new details are:

RANDOM VIBRATIONS, SPECTRAL AND WAVELET ANALYSIS, 3rd edition, D. E.
Newland, 1993, Addison Wesley Longman, ISBN 0582 215846

MECHANICAL VIBRATION ANALYSIS AND COMPUTATION, D. E. Newland, 1989, Addison
Wesley Longman, ISBN 0582 027446

The books are no longer available from Wiley, hence the
confusion. They are now published by Addison Wesley Longman whose
customer service desk in the US can be reached by phone on 617 944
3700.

If any reader still has difficulties, please email
anna.faherty@awl.co.uk who can arrange for delivery by mail.


<webmaster@wavelet.org>