Optimal Structure from Motion: Local Ambiguities and Global Estimates

Structure From Motion (SFM) refers to the problem of estimating spatial properties of a three-dimensional scene from the motion of its projection onto a two-dimensional surface, such as the retina. We present an analysis of SFM which results in algorithms that are provably convergent and provably optimal with respect to a chosen norm.In particular, we cast SFM as the minimization of a high-dimensional quadratic cost function, and show how it is possible to reduce it to the minimization of a two-dimensional function whose stationary points are in one-to-one correspondence with those of the original cost function. As a consequence, we can plot the reduced cost function and characterize the configurations of structure and motion that result in local minima. As an example, we discuss two local minima that are associated with well-known visual illusions. Knowledge of the topology of the residual in the presence of such local minima allows us to formulate minimization algorithms that, in addition to provably converge to stationary points of the original cost function, can switch between different local extrema in order to converge to the global minimum, under suitable conditions. We also offer an experimental study of the distribution of the estimation error in the presence of noise in the measurements, and characterize the sensitivity of the algorithm using the structure of Fisher's Information matrix.     

Grafted polypropylene membranes for purification of water containing sodium alkylbenzenesulfonate

The possibility of eliminating sodium alkylbenzenesulfonate (ABS) from water by reverse osmosis was examined by operating with polypropylene membranes grafted with poly(acrylic acid). Flux through the membranes decreases with increasing thickness. For 23μ thickness and 0.5 1./hr-m² flux, the rejection is 100% (operating pressure, 50 atm: concentration, 1 g ABS/l.). For 5μ thickness the rejection is lower and is related to ABS concentration: rejection increases with increasing concentration and reaches 60% for concentrations higher than 5 g ABS/l. (flux, 51./hr-m²; operating pressure, 50 atm). This is ascribed to the prevailing ultrafiltration as regards the reverse osmosis. ABS, at higher concentrations than the critical value, gives micelles whose dimensions might be compared to pores. The membrane behavior is influenced by previous treatments; when the membrane has been conditioned in NaCl solution, it is more selective and more permeable; the rejection reaches 90% with fluxes of 6.8 l./hr-m² for concentrations of 5 g ABS/1. (operating pressure, 50 atm). In order to relate the properties to the structure of membranes, we have examined them by electron microscope. The ion exchange which takes place between membrane and NaCl and between membrane and ABS has also been studied.     

Direct solution of renormalization group equations of QCD in x-space: NLO implementations at leading twist

We illustrate the implementation of a method based on the use of recursion relations in (Bjorken) x-space for the solution of the evolution equations of QCD for all the leading twist distributions. The algorithm has the advantage of being very fast. The implementation that we release is written in C and is performed to next-to-leading order in α_s.

Program summary

Title of program:evolution.cCatalogue identifier: ADUBProgram summary URL:http://cpc.cs.qub.ac.uk/summaries/ADUBProgram obtainable from: CPC Program Library, Queen's University of Belfast, N. IrelandComputer: Athlon 1800 plusOperating system under which the program has been tested: LinuxProgramming language used: CPeripherals used: Laser printerNo. of bytes in distributed program, including test data, etc.: 4559No. of lines in distributed program, including test data, etc.: 1048Distribution format: gzip fileNature of physical problem: The program provided here solves the DGLAP evolution equations to next-to-leading order α_s, for unpolarized, longitudinally polarized and transversely polarized parton distributions.Method of solution: We use a recursive method based on an expansion of the solution in powers of log(α_s(Q)/α_s(Q₀)).Typical running time: About 1 minute and 30 seconds for the unpolarized and longitudinally polarized cases and 1 minute for the transversely polarized case.     

Globally convergent autocalibration using interval analysis

We address the problem of autocalibration of a moving camera with unknown constant intrinsic parameters. Existing autocalibration techniques use numerical optimization algorithms whose convergence to the correct result cannot be guaranteed, in general. To address this problem, we have developed a method where an interval branch-and-bound method is employed for numerical minimization. Thanks to the properties of interval analysis this method converges to the global solution with mathematical certainty and arbitrary accuracy and the only input information it requires from the user are a set of point correspondences and a search interval. The cost function is based on the Huang-Faugeras constraint of the essential matrix. A recently proposed interval extension based on Bernstein polynomial forms has been investigated to speed up the search for the solution. Finally, experimental results are presented.     

Algorithm for the retrieval of columnar water vapor from hyperspectral remotely sensed data

A new algorithm for the retrieval of columnar water vapor content is presented. The proposed procedure computes the area of the H2O absorption centered about 940 nm to allow its integrated columnar abundance as well as its density at ground level to be assessed. The procedure utilizes the HITRAN 2000 database as the source of H2O cross-section spectra. Experimental results were derived from radiometrically calibrated hyperspectral images collected by the Airborne Visible-Infrared Imaging Spectrometer (AVIRIS) sensor over the Cuprite mining district in Nevada. Numerical simulations based on the MODTRAN 4 radiative transfer code were also employed for investigating the algorithm's performance. An additional empirical H2O retrieval procedure was tested by use of data gathered by the VIRS-200 imaging spectrometer.     

Extracting Logical Schema from the Web

One of the main limitations when accessing the web is the lack of explicit structure, whose presence may help in understanding data semantics. Schema for web data can be constructed at different levels, structuring a single pages or a whole site or group of sites. Here we present an approach to give a logical schema to a web-site, first defining a model for a single page, where its contents is divided into logical sections, i.e. parts of a page each collecting related information. Then, we introduce a site model in which both physical and logical links among different page sections are represented: physical are existing hyperlinks, while logical links are links between sections containing semantically related information. We show how such links can be found and classified according to their relevance, also showing how schema is used in a structure-aware browser to improve both browsing and searching.     

Atmospheric pressure photoionization mass spectrometry

Atmospheric pressure photoionization (APPI) is the last arrival in the family of atmospheric pressure ionization (API) methods to couple mass spectrometry (MS) to liquid-phase separation techniques. The basic idea was to further extend the fields of application of liquid chromatography (LC)-MS to those molecules that are not, or are poorly amenable, to electrospray (ESI) or APCI. The present review explores the literature. After a short introduction with an historical background and the premises for its development, we describe the technique, its physical principles, and the factors that affect its efficiency. The review also presents a survey of applications in different fields.     

Reduction of immunoreactivity of bovine beta-lactoglobulin upon combined physical and proteolytic treatment

Bovine beta-lactoglobulin was hydrolyzed with trypsin or chymotrypsin before, during and after treatment at 600 MPa and pH 6.8 for 10 min at 30, 37 and 44 degrees C. The extent of beta-lactoglobulin hydrolysis under pressure was noticeably higher than at atmospheric pressure, particularly when chymotrypsin was used. Addition of proteases at ambient pressure to previously pressure-treated beta-lactoglobulin gave only a modest increase in proteolysis with respect to the untreated protein. Products of enzyme hydrolysis under pressure were separated by reverse-phase HPLC, and were found to be different from those obtained at atmospheric pressure when chymotrypsin was used. The residual immunochemical reactivity of the products of combined pressure-enzyme treatment was assessed on the unresolved hydrolysates by ELISA tests using polyclonal and monoclonal antibodies, and on individual hydrolytic fractions by Western Blotting using sera of paediatric patients allergic to whey proteins in cow milk. The immunoreactivity of the whole hydrolysates was related to their content of residual intact beta-lactoglobulin, and no immunochemical reactivity was found for all the products of chymotrypsin hydrolysis under pressure. The results indicate that chymotrypsin effectively hydrolysed hydrophobic regions of beta-lactoglobulin that were transiently exposed during the pressure treatments and that were not accessible in the native protein or in the protein that had been previously pressure treated.     

The binding mode of progesterone to its receptor deduced from molecular dynamics simulations

An unambiguous understanding of the binding mode of human progesterone to its receptor still eludes experimental search. According to the X-ray structure of the ligand-binding domain, only one (O3) of the two keto groups at the ligand ends (O3 and O20) should play a role. This result is in conflict with chemical intuition and the results of site-directed mutagenesis experiments. Herein, we report classical molecular dynamics simulations that reveal the dynamic nature of the binding in solution, elucidate the reasons why X-ray studies failed to determine the role of O20, and clarify the effects of the mutations. The predictive power of the force field is ensured by the consistent introduction of a first-principles representation of the ligand.     

Preface: Volume 23, Issue 2

This volume contains the proceedings of the First International Workshop on Symbolic Model Checking (SMC'99), held in Trento, Italy, on July 6, 1999, as part of the Second Federated Logic Conference (FLoC'99).Symbolic model checking is a formal technique for the verification of finite-state concurrent systems. Symbolic model checkers (e.g. SMV, VIS) have been used to verify industrial systems, ranging from hardware to communication protocols to safety critical plants and procedures. Symbolic model checking is the core technique for several industrial verification tools, and is applied in technology transfer projects. The aim of the SMC'99 workshop is to bring together active developers and users of symbolic model checkers, compare state of the art model checking techniques (e.g. compositional reasoning, abstraction, partitioning), discuss experimental results and experience reports, and promising directions for future research.Nine contributions (out of twenty-two submissions) were selected for presentation by the program committee. The workshop program is completed by two keynote invited lectures, given by Ken McMillan (Cadence Labs, USA) on Compositional Reasoning and Abstraction, and Fabio Somenzi (University of Colorado, USA) on Symbolic State ExplorationThanks are due to a number of people who helped to organize and plan the workshop. Among this group are the Program Committee members:

Adnan Aziz	(University of Texas at Austin, USA)
Armin Biere	(Verysys)
Sergio Campos	(Federal University of Minas Gerais, Brazil)
Alessandro Cimatti	(IRST, Italy)
Edmund Clarke	(Carnegie Mellon University, USA)
Danny Geist	(IBM Haifa, Israel)
Fausto Giunchiglia	(IRST, Italy)
Orna Grumberg	(Technion, Israel)
Markus Kaltenbach	(Siemens, Germany)
Carl Pixley	(Motorola, USA)

Full-size table

The following people helped in the evaluation of the submissions: Yael Abarbanel-Vinov, Neta Aizenbud-Reshef, Shoham Ben-David, Doron Bustan, Jorge Cuellar, David Deharbe, Cindy Eisner, Ranan Fraer, Edelweis Garcez, Leonid Gluhovsky, Marcelo Glusman, Jae-Young Jang, Shmuel Katz, Sharon Keidar, Monika Maidl, Robi Malik, Anamaria Martins Moreira, Shiri Moran, Peter Päppinghaus, Marco Roveri, Peter Warkentin, Karen Yorav, Jun Yuan, Yunshan Zhu.We would like to thank IBM Haifa Research Laboratory for the financial support to SMC'99. Carola Dori, Morena Carli, Marco Roveri and Adolfo Villafiorita helped in several matters related to the local organization. Finally, we are grateful to Michael Mislove for his constat support during the preparation of this electronic volume.Alessandro Cimatti and Orna Grumberg, Guest Editors