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

The Wavelet Digest - Volume 3, Issue 17 (October 19, 1994)


Today's Editor: Wim Sweldens
                sweldens@math.sc.edu


Today's Topics:
     1. Contents: The Journal of Fourier Analysis and Applications
     2. Meeting:  Workshop course on wavelets filter banks and applications
     3. Reply:    WD 3.16 (#9) (Parallel Discrete Wavelet Transform)
     4. Meeting:  Extension of Abstract Deadline SPIE 95 Orlando
     5. Preprint: preprints from Shidong Li available
     6. Preprint: Multiresolution Analysis and Wavelets on S^2 and S^3
     7. Question: Wavelets in the methods of moments
     8. Preprint: WWW preprint page from MathSoft
     9. Course:   Wavelet Analysis, University of Massachusetts, Lowell
    10. Preprint: Nonlinear wavelet shrinkage with Bayes rules ...
    11. Meeting:  ASME 15th Conference on Mechanical Vibration and Noise

Submissions:
  E-mail to wavelet@math.sc.edu with "submit" as subject.

Subscriptions:
  E-mail to wavelet@math.sc.edu with "subscribe" as subject.
  To unsubscribe, e-mail with "unsubscribe" followed by your e-mail
  address as subject. To change address, unsubscribe and resubscribe.

Preprints, references, and back issues can be obtained from our
information servers:
  Ftp:    ftp.math.sc.edu (/pub/wavelet)
  Gopher: gopher.math.sc.edu
  WWW:    http://www.math.sc.edu/~wavelet/


Current number of subscribers: 3709


From: John J. Benedetto, University of Maryland Subject: Contents: The Journal of Fourier Analysis and Applications ANNOUNCEMENT: The first issue has just appeared. CONTENTS: Survey Article -- Self-similarity in harmonic analysis by Robert Strichartz Research Articles -- Frequency-domain bounds for nonnegative, unsharply bandlimited functions by A.J.E.M. Janssen Aliasing and Poisson summation in the sampling theory of Paley-Wiener spaces by M.G. Beatty and J.R. Higgins Generalized Rudin-Shapiro systems by George Benke Orthogonal trigonometric Schauder bases of optimal degree for C(K) by R.A. Lorenz and A.A. Sahakian SUBSCRIPTION RATES: Institutional Personal In the U.S.A. and Canada $195 $80 All other countries $227 $112 EDITORIAL BOARD: EXECUTIVE MANAGING EDITOR John J. Benedetto, U. of Maryland, College Park MANAGING EDITORS Carlos A.Berenstein, U. of Maryland, College Park Eugene B. Fabes, U. of Minnesota, Minneapolis SURVEYS EDITOR Gerald B. Folland, U. of Washington, Seattle EXECUTIVE EDITORS Luis A. Caffarelli, Institute for Advanced Study Ronald R. Coifman, Yale University Ingrid Daubechies, Princeton University Charles Fefferman, Princeton University Herbert Hauptman, Nobel Laureate, Medical Foundation of Buffalo Tom Korner, Cambridge University Henry J. Landau, AT&T Bell Labs Guido Weiss, Washington University, St. Louis ASSOCIATE EDITORS Aline Bonami, Universite' de Paris-Sud Gavin Brown, University of Adelaide, Australia Paul L. Butzer, RWTH Aachen, Germany Sun-Yung Alice Chang, U. of California at Los Angeles Albert Cohen, Universite' de Paris Dauphine, France Marie Farge, Ecole Normale Superieure, Paris, France Robert Fefferman, University of Chicago Hans G. Feichtinger, University of Vienna, Austria Michael W. Frazier, Michigan State University J. William Helton, U. of California at San Diego Stephane Jaffard, CERMA, ENPC, France A.J.E.M. Janssen, Philips Research Labs, The Netherlands Bjorn Jawerth, U. of S. Carolina, Columbia Carlos Kenig, University of Chicago Stephane Mallat, Courant Institute, New York James McClellan, Georgia Inst. of Tech Frank Natterer, U. of Muenster, Germany Andrew M. Odlyzko, AT&T Bell Labs Eric Todd Quinto, Tufts University Fulvio Ricci,Politechnico Torino, Italy Robert S. Strichartz, Dept. of Mathematics, Cornell University Martin Vetterli, U. of California at Berkeley M. Victor Wickerhauser, Washington University, St. Louis Joshua Zeevi, Technion, Israel MANUSCRIPTS SHOULD BE SUBMITTED TO: The Journal of Fourier Analysis and Applications Department of Mathematics University of Maryland College Park, MD 20742 The following should be sent: 3 copies of the manuscript, an Abstract of less than 250 words, a list of key words, and Math. Reviews Classification Numbers. The final version of the manuscript must be in some form of TEX. AIM AND SCOPE: The Journal of Fourier Analysis and Applications will publish results in Fourier analysis, as well as applicable mathematics having a significant Fourier analytic component. Appropriate manuscripts at the highest research level will be accepted for publication. Because of the extensive, intricate, and fundamental relationships between Fourier analysis and so many other subjects, select and readable surveys will be published in each issue. These surveys will include historical articles, research tutorials, and expositions of specific topics. The Journal of Fourier Analysis and Applications will provide a perspective and means for centralizing and disseminating new information from the vantage point of Fourier analysis. The breadth of Fourier analysis and diversity of its applicability require that each paper should contain a clear and motivated introduction, which is accessible to all of our readers. Applications include the following topics - Antenna theory Optics Stochastic processes Coding theory Partial differential Time-frequency analysis Crystallography equations Time series Fast algorithms Potential theory Tomography Gabor theory and Prediction theory Turbulence applications Radar applications Uncertainty principles Image processing Sampling theory Waveletpackets Mathematics of material Spectral estimation Wavelet theory and behavior Speech processing applications Number theory --- Note from the editor: This message was sent to the Wavelet Digest already on August 23rd. Because of a mistake of mine, it only appears now.
From: Gil Strang <gs@math.mit.edu> Subject: Meeting: Workshop course on wavelets filter banks and applications WORKSHOP COURSE ON WAVELETS FILTER BANKS AND APPLICATIONS at San Jose State University January 13-16, 1995 This is the first announcement of a short course on filter banks, multirate signal processing, and wavelets. By sending an email message to Gilbert Strang at MIT (( gs@math.mit.edu )) you will receive the second announcement directly. The 4-day course will be given over the holiday weekend in January, to reduce air fares and allow for sightseeing in Northern California. The workshop course will be taught by Gilbert Strang ( MIT ) and Truong Nguyen ( University of Wisconsin ). The notes will include the forthcoming textbook " Wavelets, Filter Banks, and Applications " in preliminary form - publication date is June 1995. The topics will be Digital Filters: Analysis and Design Matrix Analysis: Toeplitz Matrices and MATLAB Implementation Multirate Signal Processing: Filtering, Decimation, Polyphase Wavelet Transform: Pyramid and Cascade Algorithms Daubechies Wavelets, Orthogonal and Biorthogonal Wavelets Audio and Image Compression, Quantization Effects Transient Detection and Radar Processing Non-Destructive Evaluation, Digital Communication Perfect Reconstruction: Cosine-Modulated Filter Banks Spectral Factorization and Transmultiplexers This workshop course will NOT assume an extended background in filter banks and wavelets. It is developed from an MIT graduate course emphasizing applications of the theory. Participants are invited to suggest problems and possible applications for discussion. An email message to gs@math.mit.edu mentioning "Workshop" will bring further information.
From: lyall@dc.aaec.com (Mike Lyall) Subject: Reply: WD 3.16 (#9) (Parallel Discrete Wavelet Transform) Hi, I think I may have a simple answer for your question. The algorithm you described in Wavelet Digest is known as the tree algorithm, and uses a tree-structered filter bank to accomplish this. This can be re-drawn into an M-channel filter bank by finding the transfer function along each brach of the tree. This will create an M=J+1 channel filter bank, where J is the number of levels in the tree. For example, take a J=3 level tree. The eqivalent M=4 channel FB will have the transfer functions: A3(z) = H1(z) --> H A2(z) = H0(z) H1(z^2) --> LH A1(z) = H0(z) H0(z^2) H1(z^4) --> LLH A0(z) = H0(z) H0(z^2) H0(z^4) --> LLL where H0(z) is the lowpass analysis filter and H1(z) is the highpass analysis filter. The sequence after the arrow denotes the path through the filter bank. Each of the above filters Ai(z) are octave band filters and will perform a DWT. The downsampling rate after each of this filters is calculated by multiplying together the downsamplers found along the particular branch. For the example above, the downsampling rates would be 2,4,8,8 for A3(z) to A0(z) respectively. All of this comes from the "noble multirate identities" discussed in much of the FB literature. A good source to find this is P.P. Vaidyanathan's book "Multirate Systems and Filter Banks", Prentice-Hall, 1993. The information I just gave you is in section 11.3.3, pp. 491-496. I've done some computational analysis of these structures and the tree algorithm is more efficient than this "composite", as I call it, structure. The reason lies behind the sampling rate at which the filtering operations take place. Overall, there are about the same number of filter taps for each configuration. However, in the tree-structure, many of these operations are performed at a lower rate, thus creating a savings. With the composite structure, all operations are done at the highest rate. Hence, there is a lot of redundancy involved. Polyphase implementations eliminate redundancies that exist in both structures. In the tree algorithm, the number of computations is reduced by 1/2. I'm not sure what the polyphase composite does yet, but it still looks like the polyphase tree beats it, especially for large trees. Hope this helps you out. Michael J. Lyall lyall@dc.aaec.com Atlantic Aerospace Electronics Corp. 6404 Ivy Lane, Ste. 300 12534 Woodstock Dr. Greenbelt, MD 20770 Upper Marlboro, MD 20772 Phone: (301) 982-5202 (301) 627-3228 FAX: (301) 982-5278
From: hszu%ulysses@relay.nswc.navy.mil (Harold Szu) Subject: Meeting: Extension of Abstract Deadline SPIE 95 Orlando Extension of Abstract Deadline CONFERENCE ON WAVELET APPLICATIONS (Part of SPIE 1995 International Symposium on Aerospace/Defense Sensing and Dual-Use Photonics) 17-21 April 1995 Orlando FL Marriott Orlando World Center Abstract (by mail) due date: 15 October 1994 (Extended) (by ASCII e-mail due day 20 Oct with Abstract & Principal Author Bio. address, phone, fax etc. to lisa@mom.spie.org, and copy to btelfe@ulysses.nswc.navy.mil) Manuscript due date: 23 January 1995 (firm deadline so that proceedings will be available on site) This conference will be unique in its attention to the timely applications of the wavelet transform (WT), as evident in the on-site publication of the SPIE Vol. 2242, pp.1-985, 5-8 April 1994 Orlando Florida. This first on-site publication had attracted 212 authors contributing 94 full papers, presented in four days in 11 sequential Technical Sessions: Mathematics & Equation Solvers, Communication, Signal & Image Processing, Computer Vision, Data Compression & Recognition, Radar Processing, Earth Sciences & Remote Sensing, Industrial Applications, Image Processsing, Recognition, and Implementation, Wavelets & Neural Networks, Biomedical Applications, Acoustic Signal Detection and Recognition. We hope that it will become the definitive forum for applications in this timely new topic. This year's program will include theory and all aspects of wavelet applications and subband coding implementations. Highlights will be adaptive multiresolution analysis and time/space-scale joint representations, such as adaptive continuous WT,e.g. neuronet-wavenet, and discrete subband coding:Haar, Daubechies, biorthogonal QMF, digital chips, and analog optical cubes. This will be an excellent opportunity to present your work and help define cutting edge frontiers that may be needed for future NIST/Advanced Technology Programs, FBI/next phases, and current ARPA, ONR R&D. Dual use R/D efforts & planning in these programs will be also welcomed at conference. Your abstract should contain: title, authors and affiliations as they will appear in program, correspondence for EACH author (mailing address, phone, fax, email), an indication that you are submitting to the Wavelet Applications Conference, H. Szu, chair; preference for oral or poster presentation, 250 word abstract, brief biography (50-100 words) of principal author. Abstracts should be sent to: email: abstracts@mom.spie.org (ascii format) fax: (206) 647-1445 (SPIE) mail: AeroSense'95 SPIE, P.O. Box 10, Bellingham WA 98227-0010 shipping address: 1000 20th St., Bellingham WA 98225 (206) 676-3290 Questions may be directed to SPIE or to: Dr. Harold Szu Naval Surface Warfare Center, Code B44 Silver Spring MD 20903 (301) 394-3097 (301) 394-3923 (fax) hszu@ulysses.nswc.navy.mil --- Note from the editor: for more detailed information on this meeting see WD 3.15 #5.
From: Shidong Li (Math GS) <shidong@math.umbc.edu> Subject: Preprint: preprints from Shidong Li available The following preprints are available by anonymous ftp to "ftp.math.umbc.edu" under the subdirectory "/pub/shidong", or by request from email to "shidong@math.umd.edu" Title: General Frame Decompositions, Pseudo-duals and its Application to Weyl-Heisenberg Frames. (submitted to Numerical Functional Analysis and Optimization, Sept.'94) Author: Shidong Li Filename: pseudoframe.ps Abstract: We consider frame decompositions other than merely using the dual frame. We provide a characterization of all possible frame decompositions with $l^2$-coefficients. We show that this class of frame decompositions can be generated using the theory of pseudo frame decompositions ({\it Section 2}). We demonstrate that generating the pseudo-duals ({\it Section 2}) for a given frame amounts to find left inverses of an operator similar to the usual frame operator. A general (parametric) formula for computing the pseudo-duals is provided. The usual dual frame decomposition is shown to be a special case. A successful application of the theory to Weyl-Heisenberg (WH) frames is discussed. Besides the (usual) dual frame that preserves the translation and modulation structure, we construct a class of infinite many pseudo-duals that preserve such a structure. We discuss the issue of selecting the best localized analysis waveforms (pseudo-duals) with the concept of a ``good'' Gabor expansion. We also show constructively that there are a whole set of pseudo-duals which do not have the translation and modulation structure, yet still constitute a mathematical decomposition with a given WH frame. Title: On Pseudo Frame Decompositions and a Class of Discrete Gabor Expansions. (submitted to IEEE Trans on Signal Processing, April'94) Authors: Shidong Li and Dennis M. Healy, Jr. Filename: gabor.ps Abstract: A new approach to the study of the discrete Gabor expansion (DGE) is introduced and analyzed in detail. We show that, in general, the DGE goes well beyond the usual biorthogonal decomposition. Rather, it belongs to a more general decomposition scheme which we call {\it pseudo frame decomposition}. This takes the form, in general, \[ \forall x\in\H, \tab x=\sum_n\<x,x^*_n\>x_n, \] in which $\{x^*_n\}$ does not have to be the dual frame even if $\{x_n\}$ is a frame ({\it Section 2}). It includes the usual DGE scheme, e.g., \cite{WR}, as a special case. The standard dual frame decomposition is also a special case. Algorithms are derived to compute the ``pseudo-dual'' sequence $\{x^*_n\}$ through a matrix representation. This allows us to generate various pseudo-dual window sequences with different features by simply solving a system of linear equations, as in \cite{WR}. Title: On Dimension Invariance of Discrete Gabor Expansions. (submitted to IEEE Signal Processing Letters, July'94) Author: Shidong Li Filename: DimInvariance.ps Abstract: We reveal a dimension invariance property of the (discrete) Gabor expansion (DGE). This is an important and very useful result in applications where signal dimension is large, or signals are on-going with infinite length. As a consequence of this result, a DGE scheme developed in a given (relatively small) dimension may be applied in spaces of arbitrary large dimension by simply adjusting the number of translations accordingly.
From: ilona@igpm.rwth-aachen.de (Ilona Weinreich) Subject: Preprint: Multiresolution Analysis and Wavelets on S^2 and S^3 The following preprint is available "Multiresolution Analysis and Wavelets on S^2 and S^3" by Stephan Dahlke, Wolfgang Dahmen, Eberhard Schmitt and Ilona Weinreich Abstract: In this paper, we construct a multiresolution analysis and a wavelet basis on two specific compact manifolds. Using special charts, the problem is reduced to finding appropriate nested spaces on rectangular domains. The claim of C^1-continuity gives rise to certain boundary conditions on the rectangles. To satisfy these conditions, we use a tensor product approach in which one factor is an exponential spline. Bericht Nr. 104, Institut fuer Geometrie und Praktische Mathematik, RWTH Aachen The paper can be retrieved as a compressed postscript file using the following procedure: ftp ftp.igpm.rwth-aachen.de anonymous e-mail-adress (password) cd reports bin get igpm_104.ps.Z Ilona Weinreich Intitut fuer Geometrie und Praktische Mathematik RWTH Aachen Templergraben 55 52056 Aachen e-mail: ilona@igpm.rwth-aachen.de
From: taylorb@elec.uq.oz.au (Benjamin Taylor) Subject: Question: Wavelets in the methods of moments Hi, I'm trying to reproduce the work done by Steinberg and Leviatan in the may 1993 issue of IEEE Transactions on Antennas and Propagation.It is concerned with the utilisation of wavelets in the method of moments. My question is how do you decompose the problem into wavelets? Could anyone give me an explanation using a simple wire antenna of some arbitary length, say 6 inches? Any help or suggestions gratefully received. Thanks Benjamin Taylor
From: David Blair <dblair@mathsoft.com> Subject: Preprints: WWW preprint page from MathSoft I have put together a Web page which lists every wavelet preprint I know of that is available via anonymous ftp. Each entry in the list is linked to its ftp site so you can simply click and download. The page is under construction but presently has over 100 papers listed, most of which are in postscript or compressed postscript (.Z or .gz) format. The address of the site is http://www.mathsoft.com/wavelets.html I hope that this makes searching for information on wavelets a bit easier than it is presently; if anyone has any suggestions about additions to the page let me know. Also, there is a preprint available from Hong-Ye Gao on the subject of correctly choosing thresholds when using wavelet shrinkage in spectrum estimation: H. Y. Gao, "Choice of Thresholds for Wavelet Shrinkage Estimate of the Spectrum" This paper can be download from the above WWW site or for those with only anonymous ftp access, you can also get it at ftp.mathsoft.com in the file /pub/wavelets/wavethresh.ps --- Note from the editor: You can find a link to this page through the Wavelet Digest WWW home page. Just follow the "Some other interesting places to be" link. Special thanks to Geert Uytterhoeven (geert@cs.kuleuven.ac.be) for maintaing the Wavelet Digest WWW pages.
From: KAISERG@aspen.uml.edu Subject: Course: Wavelet Analysis, University of Massachusetts, Lowell COURSE: Wavelet Analysis, 92.552 TIME: Wednesdays 6-9 PM, January 18--May 3, 1995 PLACE: Olson Bg., Rm. 306, UMass Lowell (North Campus) INSTRUCTOR: Gerald Kaiser, Mathematical Sciences Dept. TEXTBOOK: G. Kaiser, A Friendly Guide to Wavelets, Birkhauser, 1994. Call 1-800-777-4643. MORE INFO: kaiserg@woods.uml.edu, subject=COURSE
From: Brani Vidakovic <brani@isds.Duke.EDU> Subject: Preprint: Nonlinear wavelet shrinkage with Bayes rules ... Manuscript Available. Nonlinear wavelet shrinkage with Bayes rules and Bayes factors. By Brani Vidakovic, ISDS, Duke University. Abstract. Wavelet shrinkage, the method proposed by seminal work of Donoho and Johnstone is a disarmingly simple and efficient way of de-noising data. Shrinking wavelet coefficients was proposed from several optimality criteria. The most notable are the asymptotic minimax and cross-validation criteria. In this paper a wavelet shrinkage by imposing natural properties of Bayesian models on data is proposed. The performance of methods are tested on standard Donoho-Johnstone test functions. It is on: ftp isds.duke.edu //login: anonymous ~/pub/brani/papers/WavShrinkBF.ps.Z Comments most welcome! brani@isds.duke.edu
From: den@eng.cam.ac.uk (David Newland) Subject: Meeting: ASME 15th Conference on Mechanical Vibration and Noise Call for Papers ASME 15th Biennial Conference on Mechanical Vibration and Noise Boston, Massachusetts, September 17-21, 1995 Symposium on Transient Signal Processing and Wavelets in Vibrations and Acoustics Suggested topics include but are not limited to: Application of wavelets to vibration and sound Time-frequency analysis Hilbert transform and covariance methods Envelope or phase-based processing Machine diagnostics Transient wave propagation Inverse problems and system identification Measurement techniques and signal processing algorithms Impact and shock response Isolation for transient response Important Dates: Abstract Deadline: December 15, 1994 Manuscripts (Four Copies) January 15, 1995 Notification of Acceptance April 1, 1995 Final papers on Mats April 25, 1995 Submit a one-page abstract by December 15, 1994 to either of the symposium organizers: Professor Andres Soom Mechanical and Aerospace Engineering Department State University of New York at Buffalo 321 Jarvis Hall, Buffalo, NY 14260-4400, USA Phone: (716) 645-2593, ext. 2236, Fax: (716) 645-3875 E-mail: mecsoom@ubvms.cc.buffalo.edu Papers dealing with wavelet-based methods should be sent to: Professor David E Newland Department of Engineering University of Cambridge Trumpington Street Cambridge, CB2 1PZ, UK Phone: +44 223 332670, Fax: +44 223 359153 E-mail: den@eng.cam.ac.uk Professor David Newland Cambridge University Engineering Department Trumpington St Cambridge, CB2 1PZ, UK



<webmaster@wavelet.org>