Hierarchical multiple-SIMD architecture for image analysis

Real-time image analysis requires the use of massively parallel machines. Conventional parallel machines consist of an array of identical processors organized in either single instruction multiple data (SIMD) or multiple instruction multiple data (MIMD) configurations. Machines of this type generally only operate effectively on parts of the image analysis problem. SIMD on the low level processing and MIMD on the high level processing. In this paper we describe the Warwick Pyramid Machine, an architecture consisting of both SIMD and MIMD parts in a multiple-SIMD (MSIMD) organization which can operate effectively at all levels of the image analysis problem.     

Release of endothall from aquathol granular aquatic herbicide

The release of endothall from Aquathol Granular Aquatic Herbicide (Aquathol) was determined in two static shake flask tests. A zero-order release rate of 1.58 mg l⁻¹ h⁻¹ and a half-life of 5.06 h were calculated from the replicate studies. These experiments show that Aquathol releases endothall relatively rapidly in a static environment; however, this release would be expected to be faster in a dynamic aquatic system.     

Ultra large scale static rams

Using a patented defect avoidance technique, high yield production of high density SRAM devices (ULSI SRAMs) can be achieved one process generation ahead of the rest of the industry. Production wafer yields as high as 100% and long-term average yields above 80% are reported on Inova's monolithic, 1.2μ, 320 square mm, one megabit SRAM demonstrating a practical method of achieving wafer scale integration. A yield model is presented and used to determine the optimized architecture and redundancy scheme for Inova's four megabit SRAM and to predict yield as a function of defect density. Achievement of a working 8M-bit experimental device using a 1.2μ process is also reported.     

Auswirkungen kulturspezifischer Verhaltensmuster auf das Sozialkapital in multinationalen IT-Projektteams

This paper contributes to a better understanding and to mitigate negative consequences of cultural diversity in multinational IT project teams. Our research explores how culture-specific behaviors impact social capital among team members and how firms can manage the strains. In the existing IS culture literature, culture-specific behaviors are ?C if at all ?C traced back to single culture dimensions. In contrast, the approach proposed in this article goes one step further suggesting that it is necessary to combine several culture dimensions to better understand a certain culture-specific behavior and consequently be able to better manage resulting relationship problems in multinational settings. Conducting exploratory case studies in six multinational IT projects, two exemplary cultural behavior patterns (face maintenance in India and post-communism in the Czech Republic) are identified, and management actions to avoid project performance problems are derived. The results contribute to a better understanding and management of the negative impact of culture-specific behaviors in IT project teams and corroborate that research based on culture dimensions, such as those conceptualized by Hofstede or House et al., is valuable for understanding multi-country IS projects. The findings in particular suggest that aggregating these dimensions to cultural behavior patterns improves their explanatory power and consequently the management??s capability to mitigate the negative consequences of cultural diversity.     

IPSEP-COLA: an incremental procedure for separation constraint layout of graphs

Existing information-visualization techniques that target small screens are usually limited to exploring a few hundred items. In this article we present a scatterplot tool for personal digital assistants that allows the handling of many thousands of items. The application's scalability is achieved by incorporating two alternative interaction techniques: a geometric-semantic zoom that provides smooth transition between overview and detail, and a fisheye distortion that displays the focus and context regions of the scatterplot in a single view. A user study with 24 participants was conducted to compare the usability and efficiency of both techniques when searching a book database containing 7500 items. The study was run on a pen-driven Wacom board simulating a PDA interface. While the results showed no significant difference in task-completion times, a clear majority of 20 users preferred the fisheye view over the zoom interaction. In addition, other dependent variables such as user satisfaction and subjective rating of orientation and navigation support revealed a preference for the fisheye distortion. These findings partly contradict related research and indicate that, when using a small screen, users place higher value on the ability to preserve navigational context than they do on the ease of use of a simplistic, metaphor-based interaction style     

Executing Formal Specifications with Concurrent Constraint Programming

We have implemented a technique for execution of formal, model-based specifications. The specifications we can execute are written at a level of abstraction that is close to that used in nonexecutable specifications. The specification abstractions supported by our execution technique include using quantified assertions to directly construct post-state values, and indirect definitions of post-state values (definitions that do not use equality). Our approach is based on translating specifications to the concurrent constraint programming language AKL. While there are, of course, expressible assertions that are not executable, our technique is amenable to any formal specification language based on a finite number of intrinsic types and pre- and postcondition assertions.     

Reprint of a process model for developing usable cross-cultural websites

In this paper we present a process model for developing usable cross-cultural websites. Compatible with ISO 13407, the process model documents an abstraction of the design process focusing on cultural issues in development. It provides a framework in which a variety of user-based and expert-based techniques for analysis and design are placed within the life-cycle of website development. In developing the model, we relate practical approaches to design with theories and models of culture and discuss the relevance of such theories to the practical design process. In particular we focus on four key concerns: how an audit of local website attractors can inform the design process; the concept of a cultural fingerprint to contrast websites with the cultural needs of local users; the problems associated with user evaluation; and cross-cultural team development. We then show their relation to our process model. We conclude by summarising our contribution to date within the field.     

Tracking Motion, Deformation, and Texture Using Conditionally Gaussian Processes

We present a generative model and inference algorithm for 3D nonrigid object tracking. The model, which we call G-flow, enables the joint inference of 3D position, orientation, and nonrigid deformations, as well as object texture and background texture. Optimal inference under G-flow reduces to a conditionally Gaussian stochastic filtering problem. The optimal solution to this problem reveals a new space of computer vision algorithms, of which classic approaches such as optic flow and template matching are special cases that are optimal only under special circumstances. We evaluate G-flow on the problem of tracking facial expressions and head motion in 3D from single-camera video. Previously, the lack of realistic video data with ground truth nonrigid position information has hampered the rigorous evaluation of nonrigid tracking. We introduce a practical method of obtaining such ground truth data and present a new face video data set that was created using this technique. Results on this data set show that G-flow is much more robust and accurate than current deterministic optic-flow-based approaches.     

Why Risk Assessments Fail

Over the past decades there has been a series of major corporate governance crises. After each wave post-mortems were convened and efforts made by regulators to identify root causes. The good news—or bad, depending on your perspective—for the internal audit profession is that rarely were questions raised by those commissions and regulators about the role internal audit should have played to avoid the current crisis being reviewed.

What the commissions did call for was a massive global focus on the need for boards of directors to better oversee risk in their organizations. As pressure on directors mounts globally to improve risk oversight, their dissatisfaction with traditional internal audit services is also growing. This article suggests the root cause of the mounting internal audit customer dissatisfaction globally is internal audit “paradigm paralysis”¹—a strong attachment to traditional ways of doing internal audits that no longer meet the needs of key customers. Specific recommendations are made to help internal auditors transition past the paradigm paralysis and adopt new methods that better meet the needs of its key customers. *This article first appeared in ACCA UK’s Internal Audit e-bulletin.     

COMPARING ANNs AND GENETIC PROGRAMMING FOR VOICE QUALITY ASSESSMENT POST-TREATMENT

In the U.K., the rehabilitation of a patient's voice following treatment for cancer of the larynx is managed by Speech and Language Therapists (SALT), who listen to a patient's stylized speech and then use their experience and domain knowledge to make an assessment of the current quality of the patient's voice. This process is very subjective and time consuming, and could benefit from using AI techniques to provide objective, reproducible assessments of voice quality. A comparative study of voice quality assessment post-treatment using Artificial Neural Networks (ANN), the preferred AI technique in this application area, and Genetic Programming (GP) is described, using the same dataset, training, and verification procedures. The GP approach was found to give more accurate classifications of bad quality (immediately post-treatment) and good quality (recovered) voicings than the ANN, and in addition, gave indication of the most significant parameters in the input dataset.