When computational systems become increasingly mobile and ubiquitous, digital information and the use of computational systems may increasingly be immersed into the physical and social world of objects, people and practices. However, the digital, physical and social materials that make up these hybrid spaces have different characteristics and are hard to understand for users. In addition, users are themselves part in constructing and re-constructing the hybrid spaces. The main question addressed in this thesis is whether making aspects of the digitally mediated hybrid spaces observable and accessible provides support to users. The observability may provide support for the specific task at hand or help in building an understanding for what the system does and how, an understanding that is needed to explain system output and to cope with service breakdowns. The fundament of the approach is to empower users of computational systems to actively make sense of the system themselves. Two prototype services are described, Socifer and MobiTip. Their common denominator was to make digitally mediated parts of the hybrid spaces observable to users. Without disqualifying other kinds of information, the work focussed on digitally mediated social trails of other users. Building on experience from the prototype work and an investigation into in seamful design, observability and awareness, I have investigated the effects of making a computational system’s social context observable to users in a way that - is separated from the service’s main functionality in the interface, allowing it to become peripheral and non-obtrusive; - uses simple models and little interpretation; - to some extent opens up the service to allow for user appropriation of both service content and functionality; and - is informative rather than proactive in order to empower the user rather than acting on the user’s behalf. By designing systems that fulfil these criteria I claim that the user will be supported in performing the task at hand, with or without the service, and that with service use, the user will become more and more aware of the possibilities and limitations of the underlying technology. In addition, the digitally mediated hybrid spaces where physical, social and digital contexts meet constitute application domains in themselves, domains that users may enjoy exploring.

In last years Hidden Markov Models (HMMs) have received significant attention in the task off-line handwritten text recognition (HTR). As in automatic speech recognition (ASR), HMMs are used to model the probability of an observation sequence, given its corresponding text transcription. However, in contrast to what happens in ASR, in HTR there is no standard set of local features being used by most of the proposed systems. In this thesis we propose the use of raw binary pixels as features, in conjunction with models that deal more directly with the binary data. In particular, we propose the use of Bernoulli HMMs (BHMMs), that is, conventional HMMs in which Gaussian (mixture) distributions have been replaced by Bernoulli (mixture) probability functions. The objective is twofold: on the one hand, this allows us to better modeling the binary nature of text images (foreground/background) using BHMMs. On the other hand, this guarantees that no discriminative information is filtered out during feature extraction (most HTR available datasets can be easily binarized without a relevant loss of information). In this thesis, all the HMM theory required to develop a HMM based HTR toolkit is reviewed and adapted to the case of BHMMs. Specifically, we begin by defining a simple classifier based on BHMMs with Bernoulli probability functions at the states, and we end with an embedded Bernoulli mixture HMM recognizer for continuous HTR. Regarding the binary features, we propose a simple binary feature extraction process without significant loss of information. All input images are scaled and binarized, in order to easily reinterpret them as sequences of binary feature vectors. Two extensions are proposed to this basic feature extraction method: the use of a sliding window in order to better capture the context, and a repositioning method in order to better deal with vertical distortions. Competitive results were obtained when BHMMs and proposed methods were applied to well-known HTR databases. In particular, we ranked first at the Arabic Handwriting Recognition Competition organized during the 12th International Conference on Frontiers in Handwriting Recognition (ICFHR 2010), and at the Arabic Recognition Competition: Multi-font Multi-size Digitally Represented Text organized during the 11th International Conference on Document Analysis and Recognition (ICDAR 2011). In the last part of this thesis we propose a method for training BHMM classifiers using In last years Hidden Markov Models (HMMs) have received significant attention in the task off-line handwritten text recognition (HTR). As in automatic speech recognition (ASR), HMMs are used to model the probability of an observation sequence, given its corresponding text transcription. However, in contrast to what happens in ASR, in HTR there is no standard set of local features being used by most of the proposed systems. In this thesis we propose the use of raw binary pixels as features, in conjunction with models that deal more directly with the binary data. In particular, we propose the use of Bernoulli HMMs (BHMMs), that is, conventional HMMs in which Gaussian (mixture) distributions have been replaced by Bernoulli (mixture) probability functions. The objective is twofold: on the one hand, this allows us to better modeling the binary nature of text images (foreground/background) using BHMMs. On the other hand, this guarantees that no discriminative information is filtered out during feature extraction (most HTR available datasets can be easily binarized without a relevant loss of information). In this thesis, all the HMM theory required to develop a HMM based HTR toolkit is reviewed and adapted to the case of BHMMs. Specifically, we begin by defining a simple classifier based on BHMMs with Bernoulli probability functions at the states, and we end with an embedded Bernoulli mixture HMM recognizer for continuous HTR. Regarding the binary features, we propose a simple binary feature extraction process without significant loss of information. All input images are scaled and binarized, in order to easily reinterpret them as sequences of binary feature vectors. Two extensions are proposed to this basic feature extraction method: the use of a sliding window in order to better capture the context, and a repositioning method in order to better deal with vertical distortions. Competitive results were obtained when BHMMs and proposed methods were applied to well-known HTR databases. In particular, we ranked first at the Arabic Handwriting Recognition Competition organized during the 12th International Conference on Frontiers in Handwriting Recognition (ICFHR 2010), and at the Arabic Recognition Competition: Multi-font Multi-size Digitally Represented Text organized during the 11th International Conference on Document Analysis and Recognition (ICDAR 2011). In the last part of this thesis we propose a method for training BHMM classifiers using In last years Hidden Markov Models (HMMs) have received significant attention in the task off-line handwritten text recognition (HTR). As in automatic speech recognition (ASR), HMMs are used to model the probability of an observation sequence, given its corresponding text transcription. However, in contrast to what happens in ASR, in HTR there is no standard set of local features being used by most of the proposed systems. In this thesis we propose the use of raw binary pixels as features, in conjunction with models that deal more directly with the binary data. In particular, we propose the use of Bernoulli HMMs (BHMMs), that is, conventional HMMs in which Gaussian (mixture) distributions have been replaced by Bernoulli (mixture) probability functions. The objective is twofold: on the one hand, this allows us to better modeling the binary nature of text images (foreground/background) using BHMMs. On the other hand, this guarantees that no discriminative information is filtered out during feature extraction (most HTR available datasets can be easily binarized without a relevant loss of information). In this thesis, all the HMM theory required to develop a HMM based HTR toolkit is reviewed and adapted to the case of BHMMs. Specifically, we begin by defining a simple classifier based on BHMMs with Bernoulli probability functions at the states, and we end with an embedded Bernoulli mixture HMM recognizer for continuous HTR. Regarding the binary features, we propose a simple binary feature extraction process without significant loss of information. All input images are scaled and binarized, in order to easily reinterpret them as sequences of binary feature vectors. Two extensions are proposed to this basic feature extraction method: the use of a sliding window in order to better capture the context, and a repositioning method in order to better deal with vertical distortions. Competitive results were obtained when BHMMs and proposed methods were applied to well-known HTR databases. In particular, we ranked first at the Arabic Handwriting Recognition Competition organized during the 12th International Conference on Frontiers in Handwriting Recognition (ICFHR 2010), and at the Arabic Recognition Competition: Multi-font Multi-size Digitally Represented Text organized during the 11th International Conference on Document Analysis and Recognition (ICDAR 2011). In the last part of this thesis we propose a method for training BHMM classifiers using discriminative training criteria, instead of the conventionalMaximum Likelihood Estimation (MLE). Specifically, we propose a log-linear classifier for binary data based on the BHMM classifier. Parameter estimation of this model can be carried out using discriminative training criteria for log-linear models. In particular, we show the formulae for several MMI based criteria. Finally, we prove the equivalence between both classifiers, hence, discriminative training of a BHMM classifier can be carried out by obtaining its equivalent log-linear classifier. Reported results show that discriminative BHMMs clearly outperform conventional generative BHMMs. / Giménez Pastor, A. (2014). Bernoulli HMMs for Handwritten Text Recognition [Tesis doctoral no publicada]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/37978 / TESIS

No description available.

Earthquakes are among natural disasters with significant damage potential / however it is possible to reduce the losses by taking several remedies. Reduction of seismic losses starts with identifying and estimating the expected damage to some accuracy. Since both the design styles and the construction defects exhibit mostly local properties all over the world, damage estimations should be performed at regional levels. Another important issue in disaster mitigation is to determine a robust measure of ground motion intensity parameters. As of now, well-built correlations between shaking intensity and instrumental ground motion parameters are not yet studied in detail for Turkish data. In the first part of this thesis, regional empirical Damage Probability Matrices (DPMs) are formed for Turkey. As the input data, the detailed damage database of the 17 August 1999 Kocaeli earthquake (Mw=7.4) is used. The damage probability matrices are derived for Sakarya, Bolu and Kocaeli, for both reinforced concrete and masonry buildings. Results are compared with previous similar studies and the differences are discussed. After validation with future data, these DPMs can be used in the calculation of earthquake insurance premiums. In the second part of this thesis, two relationships between the felt-intensity and peak ground motion parameters are generated using linear least-squares regression technique. The first one correlates Modified Mercalli Intensity (MMI) to Peak Ground Acceleration (PGA) whereas the latter one does the same for Peak Ground Velocity (PGV). Old damage reports and isoseismal maps are employed for deriving 92 data pairs of MMI, PGA and PGV used in the regression analyses. These local relationships can be used in the future for ShakeMap applications in rapid response and disaster management activities.

This work introduces a new UIMS (User Interface Management System), which aims to solve numerous problems in the field of user-interface development arising from hard-coded use of user interface toolkits. The presented solution is a concrete system architecture based on the abstract ARCH model consisting of an interface abstraction-layer, a dialog definition language called GIML (Generalized Interface Markup Language) and pluggable interface rendering modules. These components form an interface toolkit called GITK (Generalized Interface ToolKit). With the aid of GITK (Generalized Interface ToolKit) one can build an application, without explicitly creating a concrete end-user interface. At runtime GITK can create these interfaces as needed from the abstract specification and run them. Thereby GITK is equipping one application with many interfaces, even kinds of interfaces that did not exist when the application was written. It should be noted that this work will concentrate on providing the base infrastructure for adaptive/adaptable system, and does not aim to deliver a complete solution. This work shows that the proposed solution is a fundamental concept needed to create interfaces for everyone, which can be used everywhere and at any time. This text further discusses the impact of such technology for users and on the various aspects of software systems and their development. The targeted main audience of this work are software developers or people with strong interest in software development.

Esse trabalho tem como objetivo o estudo e análise da implantação de um elo prismático (sexto eixo no robô Scorbot-Er V), para translação do robô em atividades didáticas visando diminuir o tempo de troca do robô e das estações de trabalho em sala de aula, melhorando a visão do aluno em relação à aprendizagem na área de automação industrial e robótica tornando-a sistêmica, dando-lhe assim capacidade de integrar vários sistemas e estudar o sistema de frenagem. Feita uma análise dos ambientes onde os alunos são treinados na programação de robôs, observou-se que no desenvolvimento desta tarefa, havia duas restrições básicas: A dificuldade na programação e movimentação física do robô na troca das estações de trabalho, que levavam em torno de três horas e que não era possível a integração de CLP, IHM, inversores de freqüência, motores AC, sensor de posicionamento e fuso de esferas no ambiente de laboratório. Desenvolveu-se uma análise da precisão do sistema mecânico do sexto eixo implantado, permitindo uma visão do funcionamento do posicionamento, através do sensor acoplado no eixo de fuso esférico, o qual é acionado por um motor AC, transmitindo e convertendo o movimento circular para linear através de uma relação de polias. O resultado obtido foi à diminuição do tempo de três horas para cinco minutos na troca das estações de trabalho e uma precisão de décimos de milímetros para o posicionamento do robô e com a integração dos sistemas foi possível melhorar a aprendizagem do aluno obtendo um ganho pedagógico. / This work aims to study and analyze the implementation of a prismatic link (sixth axis in robot Scorbot-Er V) for translation of the robot in didactic activities to reduce the time to change the robot and workstations in the classroom, improving students vision for learning in the area of industrial automation and robotics becoming systemic, thus giving you ability to integrate various systems and studying the braking system. Made an analysis of the environments where students are trained in the programming of robots and it was observed that the development of this task, there were two basic restrictions: The difficulty in planning and handling of the robot in the physical exchange of workstations, which took around three hours and it was not possible the integration of PLC, HMI, frequency inverters, AC motor, sensor of positioning and axis ball screew in a laboratory environment. It was developed an analysis of the accuracy of the mechanical of sixth axis implanted, allowing a view of the positioning operation by the sensor coupled to the axis ball screew, which is powered by an AC motor transmitting and converting the linear motion to move through a relationship of pulleys. The result was a reduction of time in three hours to five minutes in the exchange of workstations and a precision of tenths of millimeters for positioning the robot and the integration of the systems have improved the learning of students achieving pedagogical gain.

Different security issues are a top subject around the world, especially since the terror threats seem to intensify. In the same time, the transport industry suffer from problems with smuggling and theft of valuable goods. One way to increase the security might be to have a verification system installed in commercial trucks, in order to assure that the driver is the proper one. This thesis has two purposes. One is to find appropriate methods for driver verification and build a prototype of a verification system which can be used for testing and further development. The other is to study how truck drivers perceive such a system and how their conception goes along with the growing demand for higher security. The present work is the result of a cooperation between an engineer and a cognitive scientist. The thesis focuses on the transport industry and was performed for Volvo Technology Corporation (VTEC), Gothenburg, Sweden. Eleven available verification methods were studied. To enable a well-based selection of methods to implement in the prototype, inquiries and interviews with truck drivers and haulage contractors were carried out to complement the theoretical study. One regular and three biometric verification methods were chosen for the test; fingerprint verification, face recognition, voice recognition and PIN verification. These methods were put together to a prototype system that was implemented in a truck simulator. A graphical user interface was developed in order to make the system user friendly. The prototype system was tested by 18 truck drivers. They were thoroughly interviewed before and after the test in order to retrieve their background, expectations and opinions as well as their perceptions and experiences of the test. Most of the test participants were positive to the prototype system. Even though they did not feel a need for it today they believed it to “be the future”. However, some participants felt uncomfortable with the system since they felt controlled by it. It became clear how important it is to have a system that respect the users’ privacy and to assure that the users are well informed about how the system is used. Some of the technology used for the verification system requires more development to fit in the automotive context, but it is considered to be possible to achieve a secure and robust system.

This work introduces a new UIMS (User Interface Management System), which aims to solve numerous problems in the field of user-interface development arising from hard-coded use of user interface toolkits. The presented solution is a concrete system architecture based on the abstract ARCH model consisting of an interface abstraction-layer, a dialog definition language called GIML (Generalized Interface Markup Language) and pluggable interface rendering modules. These components form an interface toolkit called GITK (Generalized Interface ToolKit). With the aid of GITK (Generalized Interface ToolKit) one can build an application, without explicitly creating a concrete end-user interface. At runtime GITK can create these interfaces as needed from the abstract specification and run them. Thereby GITK is equipping one application with many interfaces, even kinds of interfaces that did not exist when the application was written. It should be noted that this work will concentrate on providing the base infrastructure for adaptive/adaptable system, and does not aim to deliver a complete solution. This work shows that the proposed solution is a fundamental concept needed to create interfaces for everyone, which can be used everywhere and at any time. This text further discusses the impact of such technology for users and on the various aspects of software systems and their development. The targeted main audience of this work are software developers or people with strong interest in software development.

Photonic switches are becoming key components in advanced optical networks because of the large variety of applications that they can perform. One of the key advantages of photonic switches is that they redirect or convert light without having to make any optical to electronic conversions and vice versa, thus allowing networking functions to be lowered into the optical layer. InP-based switches are particularly attractive because of their small size, low electrical power consumption, and compatibility with integration of laser sources, photo-detectors, and electronic components. In this dissertation the development of integrated InP photonic switches using an area-selective zinc diffusion process has been investigated. The zinc diffusion process is implemented using a semi-sealed open-tube diffusion technique. The process has proven to be highly controllable and reproducible by carefully monitoring of the diffusion parameters. Using this technique, isolated p-n junctions exhibiting good I-V characteristics and breakdown voltages greater than 10 V can be selectively defined across a semiconductor wafer. A series of Mach-Zehnder interferometric (MZI) switches/modulators have been designed and fabricated. Monolithic integration of 1x2 and 2x2 MZI switches has been demonstrated. The diffusion process circumvents the need for isolation trenches, and hence optical losses can be significantly reduced. An efficient optical beam steering device based on InGaAsP multiple quantum wells is also demonstrated. The degree of lateral current spreading is easily regulated by controlling the zinc depth, allowing optimization of the injected currents. Beam steering over a 21 microns lateral distance with electrical current values as low as 12.5 mA are demonstrated. Using this principle, a reconfigurable 1x3 switch has been implemented with crosstalk levels better than -17 dB over a 50 nm wavelength range. At these low electrical current levels, uncooled and d.c. bias operation is made feasible. The use of multimode interference (MMI) structures as active devices have also been investigated. These devices operate by selective refractive index perturbation on very specific areas within the MMI structure, and this is again realized using zinc diffusion. Several variants such as a compact MMI modulator that is as short as 350 µm, a robust 2x2 photonic switch and a tunable MMI coupler have been demonstrated.

In AP 1.4 wurde ein Verfahren zur Schaltzeitprognose verkehrsabhängiger Lichtsignalanlagen entwickelt, welches auf eine Vielzahl weiterer Lichtsignalanlagen anwendbar ist. Weiterhin wurden (AP.1.4.5) im Bereich der spurgenauen Ortung, die auf Basis von GPS ermittelten Positionen durch Fusion mit anderen Sensordaten, wie der axialen Beschleunigungen und den Drehraten um die Fahrzeughochachse sowie der Einbeziehung einer digitalen Karte (Digital Enhanced Map), diese hinsichtlich einer Spurdetektion weiterhin verbessert. Im Bereich der Datenübertragung (LSA-Fzg.) konnte die erste Teilstrecke von der Verkehrsmanagementzentrale zum Serviceprovider im Labor untersucht werden. In AP 2.1 wurde eine auf der optimalen Steuerung basierte Methode zum Energiemanagement von seriellen Hybriden entwickelt. Die optimale Ansteuerung von Motor-Start-Stopp, Gangwahl und Momentenaufteilung wird modellprädiktiv unter Beachtung des Kraftstoffverbrauchs und der Schademissionen berechnet. Nach Anpassung auf praktische Randbedingungen, lässt sich diese Methode in zukünftigen Hybridfahrzeugen als optimales Energiemanagement nutzen. Die in AP 3.1 entwickelte Softwareumgebung zur gekoppelten Fahrzeug und Verkehrssimulation wurde an Beispielszenarien getestet. Für ein Modell der Versuchsstrecke wurde umfangreiche Analysen des Ampelassistenzfunktion in komplexen Verkehrsszenarien durchgeführt. Für eine Variation verschiedener Parameter, wie Wirkreichweite, Verkehrsstärke, usw. konnten Aussagen über das Potential getroffen werden. In Zusammenarbeit mit AP 3.3 wurde ein Ampelassistenzsystem und die Ansteuerung des Active-Force-Feedback Pedals im Demonstrator implementiert. In AP 3.3 wurde ein Konzept zur Darstellung von LSA-Daten im Fahrzeug erarbeitet. Dieses wurde in einem Versuchsträger umgesetzt. Dazu wurde der Versuchsträger hardwareseitig ertüchtigt, und für die Untersuchung verschiedener Varianten der Darstellung eingesetzt.