PicSOM – content-based image retrieval with self-organizing maps

We have developed a novel system for content-based image retrieval in large, unannotated databases. The system is called PicSOM, and it is based on tree structured self-organizing maps (TS-SOMs). Given a set of reference images, PicSOM is able to retrieve another set of images which are similar to the given ones. Each TS-SOM is formed with a different image feature representation like color, texture, or shape. A new technique introduced in PicSOM facilitates automatic combination of responses from multiple TS-SOMs and their hierarchical levels. This mechanism adapts to the user's preferences in selecting which images resemble each other. Thus, the mechanism implements a relevance feedback technique on content-based image retrieval. The image queries are performed through the World Wide Web and the queries are iteratively refined as the system exposes more images to the user.     

CrasyDSE: A framework for solving Dyson–Schwinger equations

Dyson–Schwinger equations are important tools for non-perturbative analyses of quantum field theories. For example, they are very useful for investigations in quantum chromodynamics and related theories. However, sometimes progress is impeded by the complexity of the equations. Thus automating parts of the calculations will certainly be helpful in future investigations. In this article we present a framework for such an automation based on a C++ code that can deal with a large number of Green functions. Since also the creation of the expressions for the integrals of the Dyson–Schwinger equations needs to be automated, we defer this task to a Mathematica notebook. We illustrate the complete workflow with an example from Yang–Mills theory coupled to a fundamental scalar field that has been investigated recently. As a second example we calculate the propagators of pure Yang–Mills theory. Our code can serve as a basis for many further investigations where the equations are too complicated to tackle by hand. It also can easily be combined with DoFun, a program for the derivation of Dyson–Schwinger equations.¹Program summaryProgram title: CrasyDSECatalogue identifier: AEMY _v1_0Program summary URL: http://cpc.cs.qub.ac.uk/summaries/AEMY_v1_0.htmlProgram obtainable from: CPC Program Library, Queen’s University, Belfast, N. IrelandLicensing provisions: Standard CPC licence, http://cpc.cs.qub.ac.uk/licence/licence.htmlNo. of lines in distributed program, including test data, etc.: 49030No. of bytes in distributed program, including test data, etc.: 303958Distribution format: tar.gzProgramming language: Mathematica 8 and higher, C++.Computer: All on which Mathematica and C++ are available.Operating system: All on which Mathematica and C++ are available (Windows, Unix, Mac OS).Classification: 11.1, 11.4, 11.5, 11.6.Nature of problem: Solve (large) systems of Dyson–Schwinger equations numerically.Solution method: Create C++ functions in Mathematica to be used for the numeric code in C++. This code uses structures to handle large numbers of Green functions.Unusual features: Provides a tool to convert Mathematica expressions into C++ expressions including conversion of function names.Running time: Depending on the complexity of the investigated system solving the equations numerically can take seconds on a desktop PC to hours on a cluster.     

Plasma modification and deposition on inner tube faces by pulsed DC discharges

The modification of inner surfaces of tubes becomes increasingly important in many fluid transport applications. However, long tubes require special plasma techniques for a surface functionalization on the nanometer scale without damaging the bulk. This work focuses on pulsed DC glow discharges in inert argon, reactive oxygen, nitrogen and air plasma as well as polymerizing acetylene plasma revealing strong influences to the wetting behavior by the total duration of the plasma discharge time (plasma on-time). While the investigated polymers polyamide, polycarbonate, polyethylene terephtalate, and thermoplastic polyurethane react very differently on the inert and reactive etching (activation) plasma, the deposition of plasma polymer films (polymer-like carbon with decreasing cross-linking at lower plasma intensity and >50% hydrogen content) suppresses all substrate influences.     

Europeanizing planning education and the enduring power of national institutions

It is widely observed that planning education in Europe is affected increasingly by various forces of Europeanisation and internationalization. While these trends are not regarded uncritically, very often commentators treat them as inexorable, or it is assumed that individual universities have considerable agency in how they respond. However, closer attention to the literature on Europeanisation shows that nation-states significantly mediate these processes. This is the focus of this paper, which draws on a case study of the creation of an integrated multi-country Masters programme in spatial planning. The analysis shows that national institutions still exert significant force, at least on the structure of the degree programmes that can be created, which universities seeking to Europeanize have to negotiate.     

A priori and a posteriori analyses of algebraic flame surface density modeling in the context of Large Eddy Simulation of turbulent premixed combustion

The flame surface density (FSD) based reaction rate closure is one of the most important methodologies of turbulent premixed flame modeling in the context of Large Eddy Simulations (LES). The transport equation for the Favre-filtered reaction progress variable needs closure of the filtered reaction rate and the subgrid scalar flux (SGSF). The SGSF in premixed turbulent flames has both gradient and countergradient components, where the former is typically modeled using eddy diffusivity and the latter can be modeled either on its own or in combination with the filtered reaction rate term using an appropriate wrinkling factor. The scope of the present work is to identify an explicit SGSF closure for the optimum performance in combination with an already established LES FSD model. The performance of different SGSF models for premixed turbulent combustion has been assessed recently by the authors using a Direct Numerical Simulation (DNS) database of freely propagating turbulent premixed flames with a range of different values of turbulent Reynolds number. The two most promising models have been implemented in the LES code. The modeling methodology identified based on a priori DNS analysis is assessed further a posteriori by comparing the LES simulation results of turbulent methane Bunsen flames with the well-documented experimental data. A significant change of the overall flame speed is not observed for different SGSF models. However, the flame shape and thickness respond to the modeling of SGSF. Considering the fact that the SGSF models have very different characteristics, the overall effect on the LES results in this work is smaller than expected. An extension of a previous a priori DNS analysis provides detailed explanations for the observed behavior.     

Core/Shell and Hollow Ultra High Molecular Weight Polyethylene Nanofibers and Nanoporous Polyethylene Prepared by Mesoscopic Shape Replication Catalysis

Polyvinyl alcohol (PVA) nanofibers, produced by electrospinning, represent attractive high‐surface‐area supports for olefin polymerization catalysts. Tethered with metal alkyls, PVA nanofibers immobilize a great variety of transition metal compounds, thus producing highly active nanofiber‐supported Ziegler‐, metallocene‐, and post‐metallocene catalysts. Whereas most conventional heterogeneous polymerization catalysts form particles, PVA‐nanofiber‐supported catalysts enable polyolefin nanofiber and nanostructure formation by mesoscopic shape replication using electrospun nanofibers as templates. At low ethylene pressure, linear correlation between average PE/PVA core/shell fiber diameter and polymerization time are made. At elevated pressure, this control is lost, accounting for the formation of reactor blends consisting of PE granules and built‐in PE/PVA nanofibers. Whereas conventional catalysts produce micrometer‐sized particles of ultrahigh molecular weight PE (UHMWPE), PVA‐nanofiber supported chromium catalysts afford new families UHMWPE materials. They range from UHMWPE/PVA core/shell nanofibers and nonwovens to hollow UHMWPE fibers and nanoporous UHMWPE, obtained by removing the PVA component.     

Comparison of different propagation steps for lattice Boltzmann methods

Several possibilities exist to implement the propagation step of lattice Boltzmann methods. This paper describes common implementations and compares the number of memory transfer operations they require per lattice node update. A performance model based on the memory bandwidth is then used to obtain an estimation of the maximum achievable performance on different machines. A subset of the discussed implementations of the propagation step are benchmarked on different Intel- and AMD-based compute nodes using the framework of an existing flow solver that is specially adapted to simulate flow in porous media, and the model is validated against the measurements. Advanced approaches for the propagation step like “A–A pattern” or “Esoteric Twist” require more programming effort but often sustain significantly better performance than non-naïve but straightforward implementations.