77 research outputs found
Effizienzprobleme weltwirtschaftlicher Organisationen
No full textSIGLEBibliothek Weltwirtschaft Kiel C 156591 / FIZ - Fachinformationszzentrum Karlsruhe / TIB - Technische InformationsbibliothekDEGerman
Evolving Networks of Expertise
No full text<p>Presentation given by Dr William Gunn during the Research Trends Virtual Seminar</p>
<p>The Individual and Scholarly Networks: Part 1: Building Networks:</p>
<p> </p>
<p>Dr William Gunn of Mendeley talked on building networks through information linking, rather than connecting people through known relationships.</p
Improving Engagement Tools to Enable Advocacy in Public Sector Activities
No full textThis research reports on the improvement of engagement activities with academics and public sector organisations in the northwest of the UK. Such activities are often supported by tools and resources public sector professionals use in their creative engagement practice. However, generic and prescribed tools do not ensure an appropriate engagement with the communities they work with. There is a need to tailor such tools for specific contexts to accommodate multiple design languages and skills of those involved in engagement processes in order to support them in design and decision-making processes. This paper presents a new design practice, where engagement practitioners develop their everyday activities through co-designing improvements of tools, enabling advocacy on the act of improvement. In this practice, engagement practitioners look at the instructions, functions, and flexibility of tools they use, provide suggestions for improvements, and make design decisions that enable better creative engagement activities through a series of co-design workshops. This research follows a Participatory Action Research (PAR) approach, where four case studies document the co-design process, and the planning and delivery of improvement workshops. The paper discusses the insights and lessons learned from the co-design process and the impact of the new improvement practice. The outcomes of this research include building design capabilities for improvement in participants, insights on the improvement of tools, and guidelines to conduct the design practice. This research contributes to the shift in design agency from ‘transferring improvements to’ to ‘co-designing improvements with’ individuals and organisations
Optimization of Hough transform for circle processing using drop technique implemented on GPGPU
No full textW artykule jest przedstawiona implementacja transformaty Hougha (HT) dla okręgów z wykorzystaniem platformy CUDA. Wykorzystano zbiór równolegle przetwarzaniach HT z synchronizacją dostępu do pamięci podręcznej tekstur. Zaproponowane rozwiązanie umożliwia przyspieszenie obliczeń o około 16% z wykorzystaniem techniki odrzucenia (dwu etapowej wyznaczania HT) względem algorytmu naiwnego. Celem pracy jest estymacja położenia i kołnierza lustra półsferycznego do pomiaru oświetlenia.Implementation of the Hough transform (HT) [1] using NVidia CUDA 4.0 platform [5, 6] is considered. The aim is fast estimation of the position and radius of a light probe. The light probe device is e.g. a hemispherical mirror (Fig. 1) used for light estimation in some points of the 3D space [3, 4] for purposes of computer animation. The GPGPU based implementation allows parallel computation of the HT. Direct implementation tests all cases, which is not computationally time feasible. Knowledge about the number of desired circles (a single circle) simplifies searching. The naive algorithm (Fig. 4) tests circles and compares temporal results with the best HT. This technique is not fast due to testing temporal HT values for every circle pixel. The proposed drop technique (Fig. 5) is based on the predefined angle threshold. The first step computes HT values up to the predefined angle threshold. The temporal result is compared with the best obtained HT value. The computations are continued (second step) if the temporal result is better than the best HT, or dropped otherwise. After the second step the comparison is calculated again. Preprocessing of the input image is necessary according to [3] and negative representation is used. The computation time depends on the number of the circle sampling points and threshold. The threshold influences the computation time. The value about 200 for the tested image is optimal. There has been obtained improvement of about 16% for the drop technique
Rheo-optical determination of the interfacial tension in a dispersed blend
No full textIn this work, we present two novel methods to determine the interfacial tension of a disperse polymer blend through rheo-optical measurements of flow-induced single drop distortions. A counter-rotating shearing device with transparent plates is used to measure drop distortions. The cell geometry allows for a top view of the deforming drop, i.e., along the velocity gradient direction. Such a view is the only possible option for all currently available commercial rheo-optical instruments. Two different quantities are monitored, namely, the drop axis along the vorticity direction, and the rotation period of the drop surface. We use drops of a polyacrilamide aqueous solution (a shear thinning liquid) immersed in a polyisobutene matrix. Experimental results are interpreted in terms of theories for Newtonian liquids, where the relevant parameter is the Capillary number. If an appropriate viscosity ratio is chosen, that accounts for the shear thinning behaviour of the drop phase, good agreement is found between measurements and theoretical predictions. As a result, a robust estimate of the blend interfacial tension, that makes use only of the information acquired from top view experiments, is obtained
Problems with Publishing: Existing State and Solutions
No full textacademic career, market for economists, publishing,
- …
