Search

Determining the right requirements to develop a system is crucial as it involved variety of people and it affects the quality of the end product. Different stakeholders have different requirements which they may express in different ways. Naturally, they will express requirements in their own terms and with implicit knowledge of their own work. Generally, stakeholders are not sure of what they want from the computer system except in the most general terms. Hence, conflicts are inevitable. Negotiation is applied to resolve the conflicts and becoming popular nowadays to better improve requirements engineering process. Therefore, this paper is empirically confirming the effectiveness of negotiation effort in order to improve the level of agreement among all the stakeholders. Agreed requirements are believed to represent all the stakeholders' perspectives and perceptions and to obtain a set of unambiguous, correct, complete, consistent and achievable requirements. [ABSTRACT FROM AUTHOR]

Software systems need to evolve, and systems built using model driven approaches are no exception. The role of the models are increasing and becoming more and more important in the software development and evolution. So, the Model-Driven Software Evolution (MoDSE) is prominent. Several stakeholders are involved in the process of evolution. These user groups (stakeholders) have different concerns relevant to the models to be evolved in MoDSE. Multiple views provide a means to visualize complex information and are also a way to fulfill the concerns of different user groups. In this paper the concept of multiple views for the MoDSE is introduced and shows how these multiple views addresses the stakeholders concerns. The different viewpoints are identified to construct the multiple views. This paper also shows an analytical support how the proposed views are closer to the Lehman's laws of software evolution. [ABSTRACT FROM AUTHOR]

This paper focuses on an empirical layout improvement project. In the implementation process of the project, the authors followed an enhanced procedure to increase the involvement of the area experts in the layout design process, which assured the successful realisation of an appropriate layout solution satisfied to both the project team and the top management of a case company. In this paper, the project, the enhanced procedure and the designed new layout for a shop floor in the case company by applying the procedure are introduced, and the summarised evaluation on the new layout and the enhanced procedure is also presented. [ABSTRACT FROM AUTHOR]

TOKAI RIKA is a manufacturer using 3D data in their operations. The company's core business lies in the production of car parts. To enhance their in-house mold-making operations, the company decided to switch from paper to 3D data. The company proved very successful by using a unique method of coloring work specifications. They are using 3D to improve engineering communication in mold design. [ABSTRACT FROM AUTHOR]

The general thermodynamically based theory of the constitutive and evolution equations of elastic-brittle damaged materials in a total stress-strain formulation, developed by Murakami and Kamiya [162], is the key for a further extension presented in the paper by Kuna-Ciska l and Skrzypek [208]. The Helmholtz free energy is a function of the elastic strain tensor εe, the second-rank damage tensor D, and another scalar damage variable β. The following Helmholtz free energy decomposition into the elastic and the damage terms is postulated as the state potential (2.20): 3.1$$ \rho \Psi (\varepsilon ^e ,D,\beta ) = \rho \Psi ^e (\varepsilon ^e ,D) + \rho \Psi ^d (\beta ) $$ [ABSTRACT FROM AUTHOR]

An enormous amount of multimedia data will be transmitted by various devices in the wireless personal area network, and this network environment will require very high transmission capacity. In this paper, we apply multiple antennas to MB-OFDM UWB system for high performance. With an emphasis on a preamble design for multichannel separation, we address the channel estimation in MB-OFDM system with multiple antennas. By properly designing each preamble for multiple antennas to be orthogonal in the time domain, the channel estimation can be applied to the MB-OFDM proposal for IEEE 802.15.3a standard in the case of more than two transmit antennas. By using the multiantenna scheme and proposed channel estimation technique, the reliability and performance of the MB-OFDM system can be improved. [ABSTRACT FROM AUTHOR]

A combination of OFDM with FDMA (called OFDMA) is regarded as a promising solution for improving the performance of BWA with multiple users. Although transmit diversity is clearly advantageous in OFDMA systems, it may not be possible that practical OFDMA systems have more than two antennas due to size, cost, or hardware limitations. Cooperative networks aim to utilize the available terminals as a collection of distributed antennas for the transmission of signals from one end of the network to the other. This paper focuses on design and performance evaluation of IEEE 802.16 MIMO-OFDMA systems with cooperative diversity. In this chapter, for high-rate and reliable transmission, efficient transmission codes of cooperative transmission are proposed for each transmission case (single transmission case or mutual transmission case). [ABSTRACT FROM AUTHOR]

Radio frequency identification (RFID) technology is not fundamentally new and concerns a whole range of applications. The first RFID application may have been the Royal British Air Force's "Identify Friend or Foe" system, which was used during the Second World War to identify friendly aircraft. RFID can be applied to a variety of tasks, structures, work systems, and contexts along the value chain, including business-to-business (B-2-B) logistics, internal operations, business-toconsumer (B-2-C) marketing, and after-sales service applications [1-6]. However, the boom that RFID technology enjoys today is basically due to the standardization [7] and development of low cost devices. Like every wireless device, RFID systems bring with them security and privacy issues to all those people who have been working in this area. Security issues involve classic attacks, namely denial of service, impersonation of tags, or channel eavesdropping. These attacks are rendered more practicable because of the tags' lack of computational and storage capacity. There are many papers investigating these issues in various ways [7, 8, 10-12, 14]. Today's challenge is to find protocols (or deployments) which allow authorized parties to identify the tags without an adversary being able to track them, thus getting to the root of the privacy problem [8, 13]. It is well known that the reason not to use well known authentication protocols is that such protocols do not preserve the privacy of the provider. Asymmetric cryptography could easily solve this problem, but it is too heavy to be implemented within a tag. [ABSTRACT FROM AUTHOR]

Searching information on the Web can be viewed as extremely easy, or it can be extremely difficult. This is because the WWW is not indexed like many library catalogs or journal-article indexes. When we search on the Web, we are not searching it directly but are actually searching the Web pages collected and indexed by a search tool from computers all over the world that contains the actual Web pages. Examples of search tools are Yahoo! Search, Google, Alta Vista, and so on. The different types of search tools have their own strengths and weaknesses. Depending on your information needs, one may work better for you than another. Search directories are hierarchical databases with references to websites. The websites that are included are classified according to the rules of that particular search service. Search engines, on the other hand, use software to crawl the Internet in search of what you would like using of terms or keywords. Specialized databases are the hidden parts of the WWW that are normally not found by regular search engines [1]. This paper presents the extension work from Integrated Filtered Web-Search Engine (IFWSE) [2] by looking at the enhancement of the program structure to retrieve the search hits. Further experiments were also carried out to compare its performance with with a well-established, meta-search engine-the MetaCrawler [3] [4]. [ABSTRACT FROM AUTHOR]

The contributions, lessons, and trends found in the DESider Symposium Proceedings are discussed. Three broad classifications can be made among the papers: The first is between studies which widen the code- and user-base of DES and its range of applications to establish the current level of quality, and those which bring out neighbouring hybrid concepts such as SAS for discussions and improvements.The second is between "natural" uses of DES with boundary layers fully covered by RANS, and uses with RANS only as a wall model in LES.The third is between single-model methods such as DES, and zonal methods even if they are closely related. [ABSTRACT FROM AUTHOR]

Hybrid RANS/LES methods combine the advantages of RANS models in terms of simulation costs, with the ability of LES to capture unsteady flow structures. A problem encountered in applications of this technique is the transmission of information between the RANS and LES zones: in the RANS region only mean-flow data is available, and the Reynolds shear stresses are entirely supported by the turbulence model; in the LES region, on the other hand, turbulent eddies must be present to support the turbulent momentum transport. The generation of these eddies is a critical ingredient in determining the effectiveness and accuracy of a hybrid method. In this paper we discuss this problem for applications in which the RANS region is below the LES zone, and in cases in which the flow advects from RANS to LES regions. We will also present recent results from our research group. [ABSTRACT FROM AUTHOR]

The paper evaluates a new method for prescribing synthesized turbulent inlet boundary conditions. When making LES, DES or hybrid LES-RANS, a precursor channel DNS is often used. The disadvantage of this method is that it is difficult to re-scale the DNS fluctuations to higher Reynolds numbers. In the present work, synthesized isotropic turbulent fluctuations are generated at the inlet plane with a prescribed turbulent length scale and energy spectrum. A large number of independent realizations are generated. A correlation in time between these realizations is introduced via an asymmetric, non-truncated time filter. In this way the turbulent time scale of the synthesized isotropic turbulent fluctuations is prescribed. [ABSTRACT FROM AUTHOR]

This paper is devoted to the study of the behaviour of DES-type methods used as Wall-Modelled-Large-Eddy-Simulation (WMLES) and to the treatment of the eddy viscosity in the framework of turbulent inlet conditions. A new method named Stimulated Detached Eddy Simulation (SDES) based on a synthetic turbulence method including pseudo-eddy viscosity synthetization is proposed and assessed on a spatially developing boundary layer. [ABSTRACT FROM AUTHOR]

The present paper summarizes the research conducted at the University of Manchester over the course of the Desider project on methods of generation of synthetic turbulence. A random procedure, referred to as the Synthetic Eddy Method (SEM), based on the classical view of turbulence as a superposition of coherent structures was developed and compared with other methods of generation of inflow boundary conditions for LES, for different types of wall bounded flows. Synthetic inflow generation methods always show a transition region downstream of the inlet until the turbulence recovers an equilibrium state. The sensitivity of this transition region to variations in the synthetic inflow turbulence energy, length scale and time scale is studied herein with applications to hybrid RANS-LES methods in mind, as RANS results cannot be expected to provide accurate values for these scales to the LES region. A parameterization of the SEM suitable for hybrid RANS-LES simulations is proposed and tested on hybrid RANS-LES simulations of several wall bounded flows, with an emphasis on the flat plat boundary layer. [ABSTRACT FROM AUTHOR]

This paper focuses on numerical aspects for hybrid RANS-LES computations using the X-LES method. In particular, the impact of using a high-order finite-volume scheme is considered. The finite-volume scheme is fourth-order accurate on non-uniform, curvilinear grids, has low numerical dispersion and dissipation, and is based on the skew-symmetric form of the compressible convection operator, which ensures that kinetic energy is conserved by convection. A limited grid convergence study is performed for the flow over a rounded bump in a square duct. The fourth-order results are shown to depend only mildly on the grid resolution. In contrast, second-order results require at least half the mesh size to become comparable to the fourth-order results. Additionally, the high-order method is extended with shock-capturing capability in such a way that interference with the subgrid-scale model is avoided. The suitability of this extension is demonstrated by means of a supersonic flow over a cavity. [ABSTRACT FROM AUTHOR]

This paper presents a first industrial attempt at Eurocopter Germany to use DES-like methods to compute the wake behind a helicopter in forward flight. Three codes have been used: TAU (DLR), CFX (Ansys) and HMB (Uni. Liverpool) and four different turbulence models: standard SST and three of its unsteady extensions SAS, DES and XLES. Results for the loads, pressure distribution and skin-friction lines have been compared to each other and against wind tunnel experimental data. These more sophisticated models brought improvements over the base SST but not always markedly and consistently: the drag gets improved although the negative lift on the body departs further form the experimental value. The pressure distribution on the backdoor has been improved by one of the computation. Oil-flow visualisation results show that the mean flow in the region of the backdoor is qualitatively reasonably well captured. [ABSTRACT FROM AUTHOR]

In this paper the three-dimensional non-reacting turbulent flow field of a swirl-stabilized gas turbine model combustor is analysed with compressible CFD. For the flow analysis URANS and Hybrid RANS/LES (DES, SAS) turbulence models were applied. The governing equations and the numerical method are described. The simulations were performed using the commercial CFD software package ANSYS CFX-10.0. The numerically achieved velocity components show a good agreement with the experimental values obtained by Particle Image Velocimetry (PIV). Furthermore, a precessing vortex core (PVC) could be found in the combustion chamber. [ABSTRACT FROM AUTHOR]

This paper presents a transonic buffet simulation on a half wing/body rigid model at a realistic Reynolds number. The modelling approach retained is a Zonal-DES approach which allows attached boundary layers to be protected and a rapid switching into LES mode in the separated areas. Transonic buffet phenomenon is a challenging case for DES type simulations since the separation point is unstable during time and the separation remains closed to the skin of the model. The simulation was performed on an adapted mesh and comparisons with experiments are carried out. It is shown that the mean field is accurately predicted with ZDES while the present RANS calculations failed to reproduce this shape of the separated area. The results also provide an insight into the unsteady nature of the flow. [ABSTRACT FROM AUTHOR]

The prediction of the laminar to turbulence transition is essential in the calculation of turbine blades, compressor blades or airfoils of airplanes since a non negligible part of the flow field is laminar or transitional. In this paper we compare the prediction capability of the Detached Eddy Simulation (DES) with the Large Eddy Simulation (LES) using the high-pass filtered (HPF) Smagorinsky model (Stolz et al., 2003) when applied to the calculation of transitional flows on turbine blades. Detailed measurements from (Canepa et al, 2003) of the well known VKI-turbine blade served to compare our results with the experiments. The calculations have been made on a fraction of the blade (10%) using non-reflective boundary conditions of Freund at the inlet and outlet plane extended to internal flows by (Magagnato et al., 2006) in combination with the Synthetic Eddy Method (SEM) proposed by (Jarrin et al., 2005). The SEM has also been extended by (Pritz et al., 2006) for compressible flows. It has been repeatedly shown that hybrid approaches can satisfactory predict flows of engineering relevance. In this work we wanted to investigate if they can also be used successfully in this difficult test case. [ABSTRACT FROM AUTHOR]

This paper presents an overview of the DES methods implemented at the Institute of Fluid Mechanics and Engineering Acoustics (ISTA) at the TU-Berlin during the course of the European DESider project. As well as the validation of these methods on the basis of simplified, academic flow cases presented in the first part, their suitability and necessity for complex industrial applications is demonstrated using results from other research projects. The methods prove robust and reliable for a wide range of applications, ranging from external to internal flows, from bluff bodies with massive separation to the partial resolution of attached boundary layers. [ABSTRACT FROM AUTHOR]

Increasingly, design studies rely on numerical tools for more and more complex situations. Unfortunately, for some, if not many, of these complex cases, steady RANS has been unable to meet the challenge and cannot capture some of the most important features. At the same time, the continuing increase in computing power has made the use of unsteady simulations feasible, although not on a day-to-day basis, for a growing number of applications. The present paper reports on the experience of implementing one such unsteady technique known as Detached Eddy Simulation (DES) in a RANS-based industrial solver. [ABSTRACT FROM AUTHOR]

The present paper presents a simulation of supersonic inlet buzz on a 20x106 points mesh, using the DDES (Delayed Detached Eddy Simulation) method. This latter approach is a version of DES that ensures the attached boundary layers to be treated in RANS. The results are compared to experimental data obtained during a previous campaign of wind-tunnel experiments. The buzz experimentally observed at Mach 1.8 is well reproduced. The buzz frequency, as well as higher frequencies existing in the experimental pressure signals, are correctly predicted. The functions of DDES prove to be robust and to perform satisfactorily. [ABSTRACT FROM AUTHOR]

The paper reports on the current state of the SST-based Scale-Adaptive Simulation model (SST-SAS model), and summarises the validation activity performed with this model in the course of the DESIDER project. The turbulence-resolving capability of the SST-SAS model has been validated with the official project test cases, covering both the underlying generic flows as well as practical aerodynamic applications. Among the latter ones are also the two aero-acoustic test cases of industrial interest. The paper includes a brief description of the current state of the SST-SAS model, as well as the illustrating test case results. Potentials of the SAS approach are briefly discussed, as well as its position amongst the other transient simulation methods. [ABSTRACT FROM AUTHOR]

This paper provides a study of the dynamic stall of a pitching NACA 0012 airfoil at 106 Reynolds number by means of numerical simulation. A 2D study is carried out comparing three OES and URANS turbulence models (URANS, URANS k-ω SST and k-ε OES). Then a 3D computation is performed using DES Spalart modelling. URANS k-ω SST is providing the best results for 2D computations. [ABSTRACT FROM AUTHOR]

This paper reports the predictions of three turbulence modelling schemes in the application to the flow around a circular cylinder in a square duct. A URANS scheme known as the Stress-Strain Lag model has been developed in the DESider project, specifically to capture the effects of stress-strain misalignment observed in unsteady mean turbulent flows (Revell, 2006). Coupling of the Lag model with the popular k -ω SST model, to form the so-called SST-Cas model, has been shown to incorporate some of the advantages of a full Reynolds Stress transport Model (RSM), whilst retaining the efficiency and stability benefits of a eddy viscosity model (EVM) (Revell et al., 2006, 2007). This paper first presents 2-D results of the cylinder case, before examining the fully 3-D case, which is also calculated using the SST-DES approach. [ABSTRACT FROM AUTHOR]

In this paper, we proposed a new subgrid scale (SGS) model in large eddy simulation (LES) of a high Reynolds number turbulent flow. The traditional approaches are usually based on modeling of the non-resolved velocity field, and are customary invariant on the Reynolds number. When the Reynolds number is high, the flow at subgrid scales is intermittent and its structure is depending on the Reynolds number. To account for this, the new SGS model is focused on simulation of the non-resolved acceleration of fluid particle. With this model, an approximation of unfiltered velocity field is reconstructed by computation of the instantaneous model-equation introduced in the paper. The performed computation of a high Reynolds number stationary homogeneous turbulence reproduced qualitatively the effects of intermittency recently observed in experimental studies. [ABSTRACT FROM AUTHOR]

Applications using rf radios operating at frequencies above 1 GHz are proliferating. The highest operating frequencies continue to increase and applications above 10 GHz and up to 77 GHz are already emerging. Systems become more complex and devices need to operate at several different frequency bands using different wireless standards. The rf-front end sections of these devices are characterized by a high diversity of components, in particular high precision passive components. In order to be produced cost-effectively, these elements need to be integrated along with the semiconductor devices. This paper describes the requirements for successful integration of rf-passive devices and proposes multilayer thin film technology as an effective rf-integration technology. [ABSTRACT FROM AUTHOR]

In automotive power applications protection and diagnosis is getting more and more important. Main reasons for this trend are safety requirements, high reliability and complex power management of modules in a car. This paper describes aspects and challenges of designing smart power high-side switches embedded in automotive systems to drive inductive, capacitive and resistive loads. [ABSTRACT FROM AUTHOR]

In this paper we will explain the trends in short-range broadband communication and investigate the technology requirements necessary for implementing the PA building block. Further we will discuss the importance of the power grid on the stability of the PA. Finally a single-ended integrated PA, part of a WiMedia/MBOA compliant UWB transceiver, is presented. [ABSTRACT FROM AUTHOR]

This paper gives an overview of several design issues for CMOS Power Amplifiers (PA) for wireless and mobile communications. The challenges, faced by the RF PA designer, and the different trade-offs, encountered during the design process, are clearly indicated. The idea of power combining is introduced and it is clarified how this technique can alleviate some of the problems related to the aggressive CMOS scaling. The theory is clarified by several design examples that cover a frequency range from the lower GHz as high as 60 GHz. [ABSTRACT FROM AUTHOR]

This paper presents a power-efficient high-speed serial interface solution, which is targeted for use in battery-operated mobile or handheld devices. Both transceiver topology and signaling protocol are described. High-speed differential and CMOS signaling are merged on the same wires. The signaling scheme contains link power management features and provides multiple communication modes to adapt efficiently to bandwidth needs. This interface has been implemented in 65nm CMOS and measurement results are shown. [ABSTRACT FROM AUTHOR]

New and improved xDSL standards are developed to bridge the last mile between the Central Office (CO) and the end user. The aDSL2+ standard doubles the bandwidth of a basic aDSL system to 2.2MHz and hence increases the bitrate up to 24Mb/s. The high Crest Factor (CF) of Discrete Multi Tone (DMT) modulated signals, however, poses serious problems on an efficient and low cost implementation of the line driver in the CO [1]. With the emerging nanometer technologies the line driver remains more than ever the bottleneck for lowering the cost and power. The low supply voltages originating from nanometer technologies increase the current density in the line driver, which affects its efficiency and reliability. This paper presents a high voltage high efficient aDSL2+ CO line driver in a mainstream nanometer CMOS technology. [ABSTRACT FROM AUTHOR]

The large scale deployment of the voice over IP, or voice over packet technology and the migration of this technology towards the edge of the network stimulates a revival of SLIC and CODEC design. The basics of this existing technology are refreshed and the architecture of a specific implementation is described in this paper. New smart power technologies with single chip low voltage CMOS and high voltage DMOS and bipolar devices allow for full single chip SLIC & CODEC integration. Another option is to put the low voltage CODEC in a deep sub micron system on chip with a mini-SLIC, integrating only the high voltage drivers. In the field new voice circuits get deployed now to bring voice to the phone over the wideband network and do the conversion from wideband IP coded voice to the analogue real world speech signals as close to the phone as possible via an Analogue Telephone Adaptor [ATA]. The conversion can be done in the phone as a real IP-phone, in a PC (Skype) via the audio of the PC or via an ATA, driving a regular phone. The conversion can also be done in the house via a VoIP to POTS converter next to the wideband modem, in your garage e.g. to drive your existing home phone cabling. The conversion can also be done with an ATA in the street cabinets driving only the last kilometer to your home phone. In this paper a Short Haul Line Integrated Circuit [SHLIC] is described, used to transmit Voice over Internet Protocol, also called VoIP, IP Telephony, Internet telephony, wideband telephony, wideband phone, and the routing of voice conversations over the Internet or through any other IP-based network. The SHLIC is a pure analogue circuit processed in a high voltage technology. Also the CODEC, performing the wideband 16Ksamples/s or A/µ law narrow band signal processing, is briefly described. [ABSTRACT FROM AUTHOR]

This paper discusses very high frequency radio links from the application level down to the circuit constraints. Because of the technical difficulties and higher cost of technologies and packaging, applications at these frequencies only make sense when the special properties of these high frequencies are offering clear advantages for these applications. Such advantages can be higher system capacity, better security and privacy, or higher spatial resolution. Exploiting these advantages requires careful choices in system design and architecture, and imposes specific constraints on circuits and technologies. In most cases, it will also require beam forming through phased array antenna structures. Implementation of the signal processing for beam forming can be achieved in an efficient way in the RF domain. [ABSTRACT FROM AUTHOR]

This paper examines the suitability of advanced SiGe BiCMOS and sub 65nm CMOS technologies for applications beyond 80GHz. System architectures are discussed along with the detailed comparison of VCOs, LNAs, PAs and static frequency dividers fabricated in CMOS and SiGe BiCMOS, as required for automotive cruise-control radar, high data-rate radio, and active and passive imaging in the 80GHz to 160GHz range. It is demonstrated experimentally that prototype SiGe HBT and BiCMOS technologies have adequate performance for all critical 80GHz building blocks, even at temperatures as high as 125 C. Although showing promise, existing 90nm GP CMOS and 65nm LP CMOS circuits at these frequencies remain significantly inferior to their SiGe counterparts. [ABSTRACT FROM AUTHOR]

Implicit in the design of a 60GHz receiver front-end is the re- quirement for broadband amplification and gain flatness in order to support advanced modulation schemes. As we have seen from lower frequency wireless systems, the choice of technology often dictates the system architecture. While either silicon CMOS or silicon Bipolar technology could be used for implementation, cost, availability, performance and time to market constraints typically dictate the technology choice. The trade-offs between CMOS and Bipolar/BiCMOS technologies for mm-wave receiver circuits are outlined in this paper. Realizing low-noise performance and gain flatness over a wide bandwidth is used as a case study from both the circuit and system points of view. [ABSTRACT FROM AUTHOR]

Summary. The problem studied here is a shortest path problem of the type encountered in sub-Riemannian geometry. It is distinguished by special structures related to its Lie group setting and the Z2 graded structure on the relevant Lie algebra. In spite of the fact that the first order necessary conditions lead to differential equations that are integrable in terms of elementary functions, in this case there remain questions related to the existence of appropriate values for the parameters which appear. In this paper we treat the problem in some generality but establish the existence of suitable parameter values only in the case of the general linear group of dimension two. [ABSTRACT FROM AUTHOR]

Summary. This paper gives a comparison principle for first-order PDEs of the Hamilton-Jacobi-Bellman type that arise in problems of nonlinear target control synthesis and reachability analysis under hard bounds on the controls. The emphasis is is on treating backward reachability sets for a system with moving target sets which may also turn to be forward reachability tubes of another system. The target sets are to be reached within preassigned intervals. The exact solutions to this problem, given in set-theoretic terms, are expressed as level sets to the solutions of some specific types of the HJB equation. But these solutions require fairly complicated calculation. The present paper presents an alternative approach that avoids exact solutions in favor of their upper and lower bounds, which in many cases may suffice for solving the required problems. For systems with linear structure ellipsoidal estimates are given, which ensure accurate approximations of nonconvex reachability sets through unions of ellipsoids. [ABSTRACT FROM AUTHOR]

Summary. During the last decade, pseudospectral (PS) optimal control methods have emerged as demonstrable efficient methods for computational nonlinear optimal control. Some fundamental problems on the feasibility and convergence of the Legendre PS method are addressed. In the first part of this paper, we summarize the main results published separately in a series of papers on these topics. Then, a new result on the feasibility and convergence is proved. Different from existing results in the literature, in this new theorem neither the invertibility of necessary conditions nor the existence of limit points is assumed. [ABSTRACT FROM AUTHOR]

Summary. In spite of the success of traditional sampled-data theory in computer control it has some disadvantages particularly for distributed, asynchronous, and multi-rate system. Event based sampling is an alternative which is explored in this paper. A simple example illustrates the differences between periodic and event based sampling. The architecture of a general structure for event based control is presented. The key elements are an event detector, an observer, and a control signal generator, which can be regarded as a generalized data-hold. Relations to nonlinear systems are discussed. Design of an event based controller is illustrated for a simple model of a micro-mechanical accelerometer. [ABSTRACT FROM AUTHOR]

Summary. We present results on omega-limit sets and minimum phase zero dynamics for hybrid dynamical systems. Moreover, we give pointers to how these results may be useful in the future for solving the output regulation problem for hybrid systems. We highlight the main attributes of omega-limit sets and we show, under mild conditions, that they are asymptotically stable. We define a minimum phase notion in terms of omega-limit sets and establish an equivalent Lyapunov characterization. Then we study the feedback stabilization problem for a class of minimum phase, relative degree one hybrid systems. Finally, we discuss output regulation for this class of hybrid systems. We illustrate the concepts with examples throughout the paper. [ABSTRACT FROM AUTHOR]

Summary. This paper is focused on the problem of output regulation for nonlinear systems within the main framework developed in [23]. The main goal is to complement that theory with some new results showing how the dimension of the internal model-based regulator can be reduced by preserving the so-called internal model property. It is shown how the problem of reducing the regulator dimension can be approached by identifying "observability" parts of the so-called steady-state input generator system. A local analysis based on canonical geometric tools and local observability decomposition is also presented to identify lower bounds on the regulator dimension. Possible benefits in designing redundant internal models are also discussed. [ABSTRACT FROM AUTHOR]

Summary. In extending the solvability of the output regulation problem to encompass more general classes of time-varying exogenous systems, various non-equivalent definitions of observability play a crucial role in the definition and properties of immersion mappings establishing the so-called "internal model property." This paper proposes a classification of the immersion mappings based on the underlying observability property, and describes the connections between different canonical realizations of internal models that fully exploit such properties for robust and adaptive output regulation design in periodic systems. [ABSTRACT FROM AUTHOR]

Summary. This paper presents a feedback method to achieve synchronization of coupled identical oscillators which are characterized by polynomial vector fields. Here, synchronization means asymptotic coincidence of the states of all the systems. Even though their models are identical, the state trajectories of the identical systems are different because of different initial conditions. Unlike other approaches where just a linear damping term is added to each system in order to achieve synchronization, we design nonlinear coupling functions between the subsystems in such a way that stability of the error dynamics between any two models results. To do that, a certain dissipation inequality and sum of squares as a computational tool are used. Finally, two examples are presented to illustrate the proposed method. [ABSTRACT FROM AUTHOR]

Summary. Up to what extent is it possible to simplify the nonlinearities of a given discrete-time system through transformations involving feedbacks or outputinjections? Solutions to these problems, at the basis of several control and state estimation design procedures, take advantage of a combined use of geometric and algebraic methods, two major setups in nonlinear control theory strongly influenced by the fundamental work of Alberto Isidori. Links between these frameworks are illustrated in the present paper which is written to celebrate his 65th birthday. [ABSTRACT FROM AUTHOR]

Summary. The paper deals with dynamic feedback linearization of multi input continuous time affine systems. The geometric properties of a dynamic feedback linearizable system as well as those of the compensator which achieves linearization are here enlightened. On the basis of these geometric properties an algorithm for the computation of a dynamic feedback obtained from the composition of regular static state feedback laws and integrators is proposed. Our result is based on the geometric approach introduced by Isidori and coworkers in 1981 for dealing with nonlinear control problems. [ABSTRACT FROM AUTHOR]

Summary. In this paper, we study the existence of periodic solutions to periodically forced systems in both the equilibrium and the nonequilibrium cases. In the equilibrium case, we prove an averaging theorem. Then, we develop the rigorous theory of the steady-state response of a nonlinear control system in order to derive some positive results on periodic forcing of autonomous systems with global Lyapunov attractors. Finally, on this occasion it is so much of a pleasure to thank Alberto Isidori for all he taught us and for all the joy we shared, and continue to share, in our research. Tanti auguri! [ABSTRACT FROM AUTHOR]

Summary. The problem of global state regulation by output feedback is investigated for a family of uncertain nonlinear systems that are bounded by a triangular system with unknown parameters. The bounding system is allowed to depend linearly on the unmeasurable states but nonlinearly on the systems output. Using the idea of universal control, together with the recent advance in nonlinear output feedback design, we develop a non-identifier based output feedback control scheme achieving, in the presence of nonlinear parameterization, global asymptotic state regulation as well as boundedness of the closed-loop system. The main contribution of the paper is two-fold: 1) the polynomial growth condition imposed in the previous work [6] is removed in this work; 2) an extension to uncertain cascade systems with zero-dynamics is carried out under mild structural and ISS conditions. [ABSTRACT FROM AUTHOR]

The purpose of this paper is to give a brief overall account of the present situation and outlook in terms of transport energy use and GHG emissions in the context of European Union transport policy. This topic is a huge one, and no attempt at comprehension is claimed. Instead the assessment will be based on a selection of a few key indicators, which the author believes are important to evaluate the present situation and identify challenges for future action and research. The indicators are all drawn from official sources or scholarly studies, while the ways in which they are interpreted and presented here are those of the author alone. During 2006 the European transport policy White Paper will undergo an official mid-term review by the European Commission. This will provide an opportunity for further reflections and possibly new directions for some of the issues to be discussed here. The present paper will first (in Section 10.2) present what we call system indicators, with the aim to highlight some critical aspects of the European transport systems' standing in terms of volume, growth, energy use and GHG emissions. Then the paper moves into policy. Section 10.3 highlights a few fundamental conditions for transport policy-making in the European Union, while Section 10.4 will go through some of the relevant targets and measures that have been adopted, evaluating them one by one with simple policy indicators. After summing up performance, Section 10.5 will address the outlooks for European transport beyond the present Kyoto agreement period, and some research issues related to policy implementation in that context will be identified. Finally a few conclusive statements are proposed, not about how much European transport, exactly, there will be room for in the common greenhouse of the future, but rather on the kinds of discussions that need to be faced in dealing with this sector's climate impact in the coming years. [ABSTRACT FROM AUTHOR]

The voluntary agreement between the European Commission and the automotive industry to cut new car CO2 emissions by 25% between 1995 and 2009 is probably the most important EU initiative to curb carbon emissions as it addresses the road transport sector, whose growth may cancel out all other attempts to meet the EU's Kyoto commitment. The paper evaluates progress in automobile fuel economy and CO2 emissions observed since 1995, examines the issue within the international context and attempts to assess whether further improvements are possible after 2010 in Europe. Based on data available up to mid-2005 it seems that, in the absence of strong technical progress and remarkable changes in consumer behavior until 2009, the industry's commitment can only be met with some years' delay. Moreover, it is most likely that technical progress will not persist in the future unless a policy mix of regulations and economic instruments is implemented. [ABSTRACT FROM AUTHOR]

The purpose of this panel is to identify implications of the papers and discussions of the past two days on the oil and gas industry. We heard that oil is being produced by unstable countries that cannot assure security of supply for a strategic commodity, and that it causes environmental degradation. However, the public needs a more balanced picture when the question of future energy supplies is being discussed. The world oil industry is meeting a huge task today. It produces around 82 million barrels a day (mn b/d), with a large part of it serving the transportation sector. There are today more than 500mn vehicles on the road worldwide. Replacing oil, or substituting other sources of energy in its place, will need more than what today's technologies can provide, and at much lower cost. Gas-to-Liquids (GTL), for example, is not forecast to grow more than 500,000 b/d by the end of decade. Moreover, demand for oil is not static. It is, in fact, rising by 1.5-2.0% annually. Global demand is forecast to reach 120mn b/d at the end of the next decade. In order to produce the 82mn b/d and to find new oil to substitute for what is being consumed and to meet future demand, the world oil industry invests around $100bn annually on exploration alone. [ABSTRACT FROM AUTHOR]

Ambient Intelligence (AmI) is considered as the composition of three emergent technologies: Ubiquitous Computing, Ubiquitous Communication and Intelligent User Interfaces. The aim of integration of aforesaid technologies is to make wider the interaction between human beings and information technology equipment through the usage of an invisible network of ubiquitous computing devices composing dynamic computational-ecosystems capable of satisfying the users' requirements. Many works focus the attention on the interaction from users to devices in order to allow an universal and immediate access to available content and services provided by the environment. This paper, vice versa, focuses on the reverse interactions, from devices to users, in order to realize a collection of autonomous control services able to minimize the human effort. In particular, thanks to a hybrid approach based on Computational Intelligence methodologies and standard Semantic Web technologies, we will describe how ubiquitous devices are able to find the suitable set of ‘intelligent' services in a transparent way. [ABSTRACT FROM AUTHOR]

The automated care and monitoring of vulnerable people is becoming a reality with the rapid development and improvement of networked sensor technologies. Many researchers have shown considerable success in determining immediate causes for concern in well-being, such as detecting falls in the home. To date however, little emphasis has been placed on monitoring well-being in a long-term sense, known as third generation telecare. The system described in this paper involves a customised sensor network, able to detect a person's movements and use of furniture and household items, coupled with a sophisticated fuzzy data analysis process able to infer activities that the monitored client is undertaking and thus answer high level queries such as "is the person eating regularly". The system also detects abnormal patterns of behaviour and provides tools for long-term trend analysis such that gradual and subtle changes in behaviour can be clearly understood. Soft computing techniques are needed due to the high degree of uncertainty in the inference process. We also describe a trial of the system, which was installed in two homes. Results indicate that fuzzy analysis enables us to summarise the data in a manner which is useful to the care providers without them needing to be experts in data analysis. [ABSTRACT FROM AUTHOR]

The aim of this paper is to discuss a fusion of the two most popular image features - color and shape - in the aspect of content-based image retrieval. It is clear that these representations have their own advantages and drawbacks. Our suggestion is to combine them to achieve better results in various areas, e.g. pattern recognition, object representation, image retrieval, by using optimal variants of particular descriptors (both, color and shape) and utilize them in the same time. To achieve such goal we propose two general strategies (sequential and parallel) for joining elementary queries. They are used to construct a system, where each image is being decomposed into regions, basing on shapes with some characteristic properties - color and its distribution. In the paper we provide an analysis of this proposition as well as the initial results of application in Content Based Image Retrieval problem. The original contribution of the presented work is related to the fusion of several shape and color descriptors and joining them into parallel or sequential structures giving considerable improvements in content-based image retrieval. The novelty is based on the fact that many existing methods (even complex ones) work in the same domain (shape or color), while the proposed approach joins features from different areas. [ABSTRACT FROM AUTHOR]

In the paper we present the employment of the direction field for contour detection and the segmentation of cylindrical objects, especially an eye iris. We propose a method of calculating the direction field for this kind of objects and present the obtained results. An analysis of the possibility of extending this approach on objects of different kinds is also included. [ABSTRACT FROM AUTHOR]

In this paper we present a robust and efficient method to overcome the negative effects of occlusion in the tracking process of multiple agents. The proposed approach is based on the matching of multiple trajectories from multiple views using spatial and temporal information. These trajectories are represented as consecutive points of a joint ground plane in the world coordinate system that belong to the same tracked agent. We introduce an integral distance between compared trajectories, which allows us to avoid mismatches, due to the possible measurement outliers in one frame. The proposed method can also be considered as an interpolation algorithm of a disconnected trajectory during the time of occlusion. This technique solves one of the most difficult problems of occlusion handling, which is a matching of two unconnected parts of the same trajectory. [ABSTRACT FROM AUTHOR]

In this paper, we evaluate the discriminative power of image features, extracted from subbands of the Gabor Wavelet Transform and the Dual-Tree Complex Wavelet Transform for the classification of zoom-endoscopy images. Further, we incorporate color channel information into the classification process and show, that this leads to superior classification results, compared to luminance-channel based image processing. [ABSTRACT FROM AUTHOR]

The paper presents some interesting properties of Grade Correspondence Analysis (GCA) used to order the pixels of gray level images described by their gray level (gl), gradient module (gm) and a family of variables dealing with pixel neighborhoods. The gm stripcharts for pixels ordered according to GCA are used to separate and then order subsets of pixels with identical descriptions of their gm neighborhoods. The spatial structure of clusters formed from subsets of adjacent clusters is then restored and used to discover compact areas of pixels with similar gm neighborhoods. Such a decomposition is performed for a MR image of a fragment of human brain. Two similar homogeneous compact areas discovered in this image suggest that the proposed method might be useful in medical classification and identification. [ABSTRACT FROM AUTHOR]

Watershed transformation is an image segmentation method. Its main advantage is that it extracts almost all edges present in the image. The edges are always continuous, which is the method's main advantage. Unfortunately it detects all edges - even if they are very weak. This causes more classes being segmented than there are objects present in the image. This phenomenon is called over-segmentation. This paper presents a solution to the over-segmentation problem. The solution uses hierarchical cluster analysis for joining similar classes of the over-segmented image into a given number of clusters. It treats watersheds as objects characterized by a set of attributes. Four different clustering methods, seven attributes combined in different sets, seven similarity/dissimilarity measures and four standardization methods were compared using quality assessment functions. Tests were performed on CT and MRI scans. Combinations that achieved the highest number of good segmentations are described. Selected segmentations are presented. [ABSTRACT FROM AUTHOR]

In this paper a novel approach for image denoising using stochastic differential equations (SDEs) is presented. In proposed method a controlled parameter to Euler's approximations of solutions to SDEs with reflecting boundary is added. It is shown that modified diffusion gives very good results for Gaussian noise source models and compares favourably with other image denoising filters. [ABSTRACT FROM AUTHOR]

The paper presents exact 3D multiview representation of monotonous polyhedrons (on the basis of their Brep) obtained by means of faces surrounding scanning algorithm. The algorithm uses the viewing sphere and perspective projection to describe the movement of points. (called K-M view space model). Visual identification is done based on these multiview models and a scene depth map. The results of tests for this algorithm for some solids are also presented. [ABSTRACT FROM AUTHOR]

Colorization is a term used to describe a computerized process for adding color to black and white pictures, movies or TV programs by replacing a scalar value stored at each pixel of the gray scale image by a vector in a three dimensional color space with luminance, saturation and hue or simply RGB. Since different colors may carry the same luminance value but vary in hue and/or saturation, the problem of colorization has no inherently "correct" solution. Due to these ambiguities, human interaction usually plays a large role. In this paper we present a novel colorization method that utilizes the morphological distance transformation and changes of neighboring pixel intensities to propagate the color within the gray scale image. The proposed method frees the user of segmenting the image, as color is provided simply by scribbles which are next automatically propagated within the image. The effectiveness of the algorithm allows the user to work interactively and to obtain the desired results promptly after providing the color scribbles. In the paper we show that the proposed method allows for high quality colorization results for still images. [ABSTRACT FROM AUTHOR]

The precise face and eyes detection is crucial in many Human-Machine Interface system. The important issue is the reliable object detection method. In this paper we present the architecture of a 3-stage face and eye detection system based on the Haar Cascade Classifiers. By applying the proposed system to the set of 10000 test images the 94% of the eyes were properly detected and precisely localized. [ABSTRACT FROM AUTHOR]

Mathematical morphology is a widely known image processing technique. Morphology-based processing techniques can be easily implemented, thanks to conceptual simplicity of basic morphological operators. However, most morphological algorithms that are run on sequential processors are unable to meet real-time requirements. A way to solve this problem are the implementations using dedicated hardware, e.g. the FPGAs. This paper describes a complete FPGA architecture for real-time edge detection. The architecture consists of prefiltering, gradient image computation and non-maximum suppression blocks, allowing the extraction of a single-pixel-thin binary edge image of good quality. The described solution, although simple and compact, is capable of processing image data stream with the speed of hundreds of frames per second for a 512×512, 8-bit gray-scale image, while keeping the costs, size and power consumption at a low level. Description of implemented algorithm, resource usage summary, output images, as well as block diagrams of the architecture are included in the paper. [ABSTRACT FROM AUTHOR]

In the paper a new segmentation method is proposed. Adaptive potential active contour approach (APAC) is a result of the relationship between active contour methods and classifier construction techniques. As an optimization algorithm simulated annealing was used to avoid local minima of energy function. The presented approach allows to obtain contours of various topology and a new adaptation mechanism can be introduced to improve segmentation results. [ABSTRACT FROM AUTHOR]

In this paper we propose a new method of generation 3D multiview representation of monotonous polyhedrons on the base of their Brep. The algorithm acts according to viewing sphere with perspective projection concept (called K-M view space model), scanning polyhedron faces in spiral way. Obtained models are used for visual identification based on them and a scene depth map. The results of algorithm testing are presented. [ABSTRACT FROM AUTHOR]

Face detection is a primary step in many applications such as face recognition, video surveillance, human computer interface, and expression recognition. Many existing detection techniques suffer under scale variation, pose variation (frontal vs. profile), illumination changes, and complex backgrounds. In this paper, we present a robust and efficient method for face detection in color images. Skin color segmentation and edge detection are employed to separate all non-face regions from the candidate faces. Primitive shape features are then used to decide which of the candidate regions actually correspond to a face. The advantage of this method is its ability to achieve a high detection rate under varying conditions (pose, scale,...) with low computational cost. [ABSTRACT FROM AUTHOR]

In the paper we introduce the method of automatic merging of 3D attribute meshes. Attributes describe the certainty of particular vertices/faces in the mesh. We calculate values of attributes on the basis of the assumed observation strategy as well as the properties of the scanning system. We show the resultant meshes and compare the method to the existing solutions. [ABSTRACT FROM AUTHOR]

In this paper, we present a new implementable learning algorithm for the general nonlinear binary classification problem. The suggested algorithm abides the maximum margin philosophy, and learns a decision function from the set of all finite linear combinations of continuous differentiable basis functions. This enables the use of a much more flexible function class than the one usually employed by Mercer-restricted kernel machines. Experiments on 2-dimensional randomly generated data are given to compare the algorithm to a Support Vector Machine. While the performances are comparable in case of Gaussian basis functions and static feature vectors the algorithm opens a novel way to hitherto intractable problems. This includes especially classification of feature vector streams, or features with dynamically varying dimensions as such in DNA analysis, natural speech or motion image recognition. [ABSTRACT FROM AUTHOR]

The problem of estimation of the preference relation in a finite set on the basis of medians from multiple pairwise comparisons, in the form of difference of ranks with random errors, with the use of nearest adjoining order idea (NAO), is investigated in the paper. The estimated form of the relation is obtained on the basis of a discrete programming problem; some probabilistic properties of the estimator are determined. They are obtained under weak assumptions about distributions of comparison errors; in particular, the expected value of comparison error of each pair can be different than zero. [ABSTRACT FROM AUTHOR]

The paper presents five algorithms for minimum vertex cover in a hypergraph. We consider: exact-backtracking, LJC, SBT, RSBT and MSBT algorithms. For a set of randomly generated hypergraphs tests have been performed to compare the aforementioned procedures. [ABSTRACT FROM AUTHOR]

This paper presents an approach based on graph mining techniques to searching for frequent patterns in design systems. The motivation for such approach is briefly presented. The graphs-based representation is used as an underlying structure for design representation. Different types of graphs used to represent designs are presented. Especially hierarchical labelled graphs are described. A possibility of adapting graph mining methods for design problems is also shown. [ABSTRACT FROM AUTHOR]

In the paper we present the method of the generalisation of a language sample for grammatical inference of quasi context-sensitive GDPLL(k) grammars. GDPLL(k) grammars and parsers have been developed as an efficient tool for syntactic pattern recognition: the grammars are characterised by very good discriminative properties and the parser for the grammars is of the linear computational complexity. Nevertheless, one of the main problems of practical application of GDPLL(k) grammars in syntactic pattern recognition systems consists in difficulties in defining the grammar from the sample of a pattern language. The method which we describe in the paper is an important element of the solution of this problem. [ABSTRACT FROM AUTHOR]

In this paper we compare two approaches to the estimation of parameters describing linear classifiers considering a feature selection possibility for multidimensional data. The parameters are determined by minimization of convex and piecewise linear penalty functionals. Optimization is performed using a base exchange strategies. We assess numerical efficiency, classification accuracy, and the dimension of the subspaces containing the found solutions. [ABSTRACT FROM AUTHOR]

In this paper, we present the hierarchical chromosome based genetic algorithm, which is used in design systems to solve complex design problems. The main goal of our work is creating of the markov chain model for the algorithm. The model contains a method of coding of the artifacts, represented by different length trees, by means of fixed length binary strings and a method of the hierarchical crossover modeling. Presented formalization is used for finding the transition matrix and the investigation of ergodicity and asymptotic properties of genetic algorithms. It also follows from the model that each population can be found with the probability equal to one within the finite number of steps. [ABSTRACT FROM AUTHOR]

In the paper a new method for binary shape characterization is proposed. It is based on the analysis of binary image pixels belonging to the internal boundary of an object performed by means of the mathematical morphology. By using morphological operators, internal boundary pixels are classified into three groups. This classification is performed for increasing sizes of structuring elements used. The changes within the class assignments are described by the boundary class distribution functions. These functions may be used as features in the pattern recognition process. [ABSTRACT FROM AUTHOR]

In syntactic pattern recognition a pattern is represented by abstract data, for instance a graph. The problem of recognition is to determine if a pattern, represented by the describing graph, belongs to a language L(G), generated by a graph grammar G. The so-called IE graphs can be used for a pattern description. They are generated by so-called ETPL(k) graph grammars. The purpose of this paper is to present an idea of a new approach to syntactic recognition of fuzzy patterns represented by fuzzy IE graphs, followed the example of random IE graphs. This methodology can be used for analysis of wider class of patterns and scenes than the one described by the classical syntactic methods. In this paper, apart from the presentation of the fuzzy-aided approach, it is also shown that the probabilistic-syntactic approach is a special case of the presented one. [ABSTRACT FROM AUTHOR]

Boosting is one of the most popular methods of multiple classification. In the paper we propose a method for speeding up the learning process by modifying the backpropagation algorithm and fuzzy clustering algorithm for boosting learning of several neuro-fuzzy classifiers. Simulations show superior performance of our method comparing to a standard boosting classification. [ABSTRACT FROM AUTHOR]

In this paper, it is shown that active hypercontours can be interpreted as a method for information fusion. The fusion concerns using of information from diverse sources and of different type (data and knowledge). Moreover, active hypercontours can work as group classifiers. [ABSTRACT FROM AUTHOR]

In this paper a method for recognition of handwritten characters based on Gabor filters and Zernike moments is discussed. The technique can be summarized as follows: in the preprocessing stage the input image is first normalized. Next the feature extraction is performed. Thus, the image is represented by the feature vector. Mean and standard deviation of the magnitude of the Gabor transform coefficients and the Zernike moments are used as features. Classification is then carried out on the basis of the extracted features. [ABSTRACT FROM AUTHOR]

The paper deals with the multistage recognition task. In this problem of recognition the Bayesian statistic is applied. The decision rules minimize the mean risk, that is the mean value of the zero-one loss function. The information on objects features is fuzzy or non-fuzzy. The probability of misclassification for local optimal strategy and the difference between probability of misclassification for the both information's are presented. Simple example of this difference conclude the work. [ABSTRACT FROM AUTHOR]

The paper presents idea of cost sensitive learning method for decision tree induction with fixed attribute acquisition cost limit. Properties of mentioned concept are established during computer experiments conducted on chosen databases from UCI Machine Learning Repository. [ABSTRACT FROM AUTHOR]

In this paper the authors presented the results of experiments on evaluation of efficiency of fast adaptive Fourier transform algorithm applied to the tasks of Fourier descriptor based contour classification. The experiments involved popular SQUID boundaries database of sea fish of different species. For object classification a nonparametrised nearest-neighbor classifier was used and distances between feature vectors were expressed in Euclidean metric. The results received for fast adaptive Fourier transform algorithm indicate significant acceleration of contour descriptors calculation and averagely more than threefold reduction in amount of contour data obtained without loss of classification accuracy. [ABSTRACT FROM AUTHOR]

A misclassification rate is most often used as a feature selection criterion. However, in the cases, when the numerical force of the training set is not sufficiently large in relation to the number of features, the risk of choosing the noisy features is very high. It produces difference between error estimations derived on the basis of the training and testing sets, so the error rate estimation can not be sufficiently confident. Feature preselection based on analysis of dependences between features is recommended in such types of tasks. An advantage of this approach is shown in the paper. As a feature selection criterion the Pearson chi-square statistics has been used. [ABSTRACT FROM AUTHOR]

This paper is about the visual features extraction. The shape visual features and appearance based visual features were used in this work for the speaker identification. A Gaussian Mixture Models (GMMs) were trained from these visual features and the speakers were recognized with the help of these GMMs. A difference between the speaker identification with GMM trained from audio or visual features was investigated too. [ABSTRACT FROM AUTHOR]

In this paper we propose different approaches to the LM integrated in a Continuous Speech Recognition system. All of them are based on classes that are made up of phrases or segments of words. The proposed models were evaluated in terms of Word Error Rate over a spontaneous dialogue corpus in Spanish. The experiments carried out show that better performance of the CSR system can be achieved introducing segments of words into a class-based LM. [ABSTRACT FROM AUTHOR]

We report in this paper on significant improvements that we have been including in our HMM-based system for recognition of ultra low resolution, anti-aliased text with small point sizes such as those frequently found in web images. First we are proposing a fully automatic training procedure where no a priori knowledge of font metrics is needed. This means that our system can be potentially built to recognize any font. Second, the system's performance can be boosted by using mixtures of Gaussians to model the probability density functions of the HMMs. Third, we show that these improvements allow the system to handle large vocabulary size up to 60'000 words with few degradation of the accuracy. We report on these results for 2 different font families, namely a serif and a sans serif font. We also report on different HMMs topologies and conclude on the benefits of using minimum duration to model the characters composing the words. [ABSTRACT FROM AUTHOR]

Statistical Machine Translation (SMT) is an evolving field where many techniques in Syntactic Pattern Recognition (SPR) are needed and applied. A typical phrase-based SMT system for translating from a T (target) language to an S (source) language contains one or more n-gram language models (LMs) and one or more phrase translation models (TMs). These LMs and TMs have a large memory footprint (up to several gigabytes). This paper describes novel techniques for filtering these models that ensure only relevant patterns in the LMs and TMs are loaded during translation. In experiments on a large Chinese-English task, these techniques yielded significant reductions in the amount of information loaded during translation: up to 58% reduction for LMs, and up to 75% for TMs. [ABSTRACT FROM AUTHOR]

In the present paper the results of applying the Elastic Graph Matching (EGM) method to classify the 23 postures occurring in Polish finger alphabet are presented. Four different classifiers were constructed. Three of them used different feature vectors to the graph's nodes description. The forth one integrated the former three classifiers. [ABSTRACT FROM AUTHOR]