We present a systematic study on gender classification with automatically detected and aligned faces. We experimented with 120 combinations of automatic face detection, face alignment and gender classification. One of the findings was that the automatic face alignment methods did not increase the gender classification rates. However, manual alignment increased classification rates a little, which suggests that automatic alignment would be useful when the alignment methods are further improved. We also found that the gender classification methods performed almost equally well with different input image sizes. In any case, the best classification rate was achieved with a support vector machine. A neural network and Adaboost achieved almost as good classification rates as the support vector machine and could be used in applications where classification speed is considered more important than the best possible classification accuracy.

No abstract

The automatic synthesis of state machines describing the behaviour of a class of objects in object‐oriented software modelling is studied. It is shown that the synthesis can be carried out on the basis of trace diagrams giving possible sequences of events during the execution of the system. An algorithm originally developed for the automatic construction of programs on the basis of their execution traces is applied to the problem, and an experimental state machine synthesizer is implemented. It is demonstrated that such a synthesizer is a highly useful component in a practical object‐oriented CASE system.

This thesis experiments with a novel approach to applying genetic algorithms in software architecture design by giving the structure of an architecture at a highly abstract level.Previously in the literature, genetic algorithms are used only to improve existing architectures. The structure and evaluation of software architectures and the principles of meta-heuristic search algorithms are introduced to give a basis to understand the implementation. Current research in the field of search-based software engineering is explored to give a perspective to the implementation presented in this thesis. The chosen genetic construction of software architectures is based on a model which contains information of a set of responsibilities and dependencies between them. An implementation using this model is presented, as well as test results achieved from a case study made on a sketch of an electronic home control system. The test results show that quality results can be achieved using the selected approach and that the presented implementation is a good starting point for future research.

No abstract

We consider the backgrounds, applications, implementations, and user interfaces of the reorderable matrix originally introduced by Jacques Bertin. As a new tool for handling the matrix, we propose a new kind of interface for interactive cluster analysis. As the main tool to order the rows and columns, we use the well-known barycenter heuristic. Two user tests are performed to verify the usefulness of the automatic tools. Information Visualization (2005) 4, 32-48.

Background and Aims: simulators are widely used in occupations where practice in authentic environments would involve high human or economic risks. surgical procedures can be simulated by increasingly complex and expensive techniques. this review gives an update on computer-based virtual reality (vr) simulators in training for laparoscopic cholecystectomies.Materials and Methods: from leading databases (medline, cochrane, embase), randomised or controlled trials and the latest systematic reviews were systematically searched and reviewed. twelve randomised trials involving simulators were identified and analysed, as well as four controlled studies. furthermore, seven studies comparing black boxes and simulators were included.Results: the results indicated any kind of simulator training (black box, vr) to be beneficial at novice level. after vr training, novice surgeons seemed to be able to perform their first live cholecystectomies with fewer errors, and in one trial the positive effect remained during the first ten cholecystectomies. no clinical follow-up data were found. optimal learning requires skills training to be conducted as part of a systematic training program. no data on the cost-benefit of simulators were found, the price of a vr simulator begins at eur 60 000.Conclusions: theoretical background to learning and limited research data support the use of simulators in the early phases of surgical training. the cost of buying and using simulators is justified if the risk of injuries and complications to patients can be reduced. Developing surgical skills requires repeated training. in order to achieve optimal learning a validated training program is needed.

No abstract