05 Fakultät Informatik, Elektrotechnik und Informationstechnik
Permanent URI for this collectionhttps://elib.uni-stuttgart.de/handle/11682/6
Browse
254 results
Search Results
Item Open Access Improving usability of gaze and voice based text entry systems(2023) Sengupta, Korok; Staab, Steffen (Prof. Dr.)Item Open Access Computational modelling of coreference and bridging resolution(2019) Rösiger, Ina; Kuhn, Jonas (Prof. Dr.)Item Open Access Rigorous compilation for near-term quantum computers(2024) Brandhofer, Sebastian; Polian, Ilia (Prof.)Quantum computing promises an exponential speedup for computational problems in material sciences, cryptography and drug design that are infeasible to resolve by traditional classical systems. As quantum computing technology matures, larger and more complex quantum states can be prepared on a quantum computer, enabling the resolution of larger problem instances, e.g. breaking larger cryptographic keys or modelling larger molecules accurately for the exploration of novel drugs. Near-term quantum computers, however, are characterized by large error rates, a relatively low number of qubits and a low connectivity between qubits. These characteristics impose strict requirements on the structure of quantum computations that must be incorporated by compilation methods targeting near-term quantum computers in order to ensure compatibility and yield highly accurate results. Rigorous compilation methods have been explored for addressing these requirements as they exactly explore the solution space and thus yield a quantum computation that is optimal with respect to the incorporated requirements. However, previous rigorous compilation methods demonstrate limited applicability and typically focus on one aspect of the imposed requirements, i.e. reducing the duration or the number of swap gates in a quantum computation. In this work, opportunities for improving near-term quantum computations through compilation are explored first. These compilation opportunities are included in rigorous compilation methods to investigate each aspect of the imposed requirements, i.e. the number of qubits, connectivity of qubits, duration and incurred errors. The developed rigorous compilation methods are then evaluated with respect to their ability to enable quantum computations that are otherwise not accessible with near-term quantum technology. Experimental results demonstrate the ability of the developed rigorous compilation methods to extend the computational reach of near-term quantum computers by generating quantum computations with a reduced requirement on the number and connectivity of qubits as well as reducing the duration and incurred errors of performed quantum computations. Furthermore, the developed rigorous compilation methods extend their applicability to quantum circuit partitioning, qubit reuse and the translation between quantum computations generated for distinct quantum technologies. Specifically, a developed rigorous compilation method exploiting the structure of a quantum computation to reuse qubits at runtime yielded a reduction in the required number of qubits of up to 5x and result error by up to 33%. The developed quantum circuit partitioning method optimally distributes a quantum computation to distinct separate partitions, reducing the required number of qubits by 40% and the cost of partitioning by 41% on average. Furthermore, a rigorous compilation method was developed for quantum computers based on neutral atoms that combines swap gate insertions and topology changes to reduce the impact of limited qubit connectivity on the quantum computation duration by up to 58% and on the result fidelity by up to 29%. Finally, the developed quantum circuit adaptation method enables to translate between distinct quantum technologies while considering heterogeneous computational primitives with distinct characteristics to reduce the idle time of qubits by up to 87% and the result fidelity by up to 40%.Item Open Access Design for reliability in advanced technologies using machine learning(2024) Klemme, Florian; Amrouch, Hussam (Prof. Dr.-Ing.)This thesis focuses on the standard cell library, which is one of the core entities in the digital circuit design flow, to demonstrate the challenges and opportunities of advanced technology nodes. The standard cell library serves as a technology interface between the foundry and the circuit designer, enabling automatic mapping of high-level circuit descriptions to the technology of the foundry through the process of logic synthesis. In the past decade, the standard cell library has been continuously adapted to keep up with the demands of shrinking process nodes. This includes, e.g., the integration of more accurate timing models, process variation, or signal integrity for cross-talk and noise in the circuit. This thesis takes this development to the next level and presents approaches to bring machine learning and transistor self-heating into the standard cell library.Item Open Access Interacting with large high-resolution display workplaces(2018) Lischke, Lars; Schmidt, Albrecht (Prof.)Large visual spaces provide a unique opportunity to communicate large and complex pieces of information; hence, they have been used for hundreds of years for varied content including maps, public notifications and artwork. Understanding and evaluating complex information will become a fundamental part of any office work. Large high-resolution displays (LHRDs) have the potential to further enhance the traditional advantages of large visual spaces and combine them with modern computing technology, thus becoming an essential tool for understanding and communicating data in future office environments. For successful deployment of LHRDs in office environments, well-suited interaction concepts are required. In this thesis, we build an understanding of how concepts for interaction with LHRDs in office environments could be designed. From the human-computer interaction (HCI) perspective three aspects are fundamental: (1) The way humans perceive and react to large visual spaces is essential for interaction with content displayed on LHRDs. (2) LHRDs require adequate input techniques. (3) The actual content requires well-designed graphical user interfaces (GUIs) and suitable input techniques. Perceptions influence how users can perform input on LHRD setups, which sets boundaries for the design of GUIs for LHRDs. Furthermore, the input technique has to be reflected in the design of the GUI. To understand how humans perceive and react to large visual information on LHRDs, we have focused on the influence of visual resolution and physical space. We show that increased visual resolution has an effect on the perceived media quality and the perceived effort and that humans can overview large visual spaces without being overwhelmed. When the display is wider than 2 m users perceive higher physical effort. When multiple users share an LHRD, they change their movement behavior depending whether a task is collaborative or competitive. For building LHRDs consideration must be given to the increased complexity of higher resolutions and physically large displays. Lower screen resolutions provide enough display quality to work efficiently, while larger physical spaces enable users to overview more content without being overwhelmed. To enhance user input on LHRDs in order to interact with large information pieces, we built working prototypes and analyzed their performance in controlled lab studies. We showed that eye-tracking based manual and gaze input cascaded (MAGIC) pointing can enhance target pointing to distant targets. MAGIC pointing is particularly beneficial when the interaction involves visual searches between pointing to targets. We contributed two gesture sets for mid-air interaction with window managers on LHRDs and found that gesture elicitation for an LHRD was not affected by legacy bias. We compared shared user input on an LHRD with personal tablets, which also functioned as a private working space, to collaborative data exploration using one input device together for interacting with an LHRD. The results showed that input with personal tablets lowered the perceived workload. Finally, we showed that variable movement resistance feedback enhanced one-dimensional data input when no visual input feedback was provided. We concluded that context-aware input techniques enhance the interaction with content displayed on an LHRD so it is essential to provide focus for the visual content and guidance for the user while performing input. To understand user expectations of working with LHRDs we prototyped with potential users how an LHRD work environment could be designed focusing on the physical screen alignment and the placement of content on the display. Based on previous work, we implemented novel alignment techniques for window management on LHRDs and compared them in a user study. The results show that users prefer techniques, that enhance the interaction without breaking well-known desktop GUI concepts. Finally, we provided the example of how an application for browsing scientific publications can benefit from extended display space. Overall, we show that GUIs for LHRDs should support the user more strongly than GUIs for smaller displays to arrange content meaningful or manage and understand large data sets, without breaking well-known GUI-metaphors. In conclusion, this thesis adopts a holistic approach to interaction with LHRDs in office environments. Based on enhanced knowledge about user perception of large visual spaces, we discuss novel input techniques for advanced user input on LHRDs. Furthermore, we present guidelines for designing future GUIs for LHRDs. Our work creates the design space of LHRD workplaces and identifies challenges and opportunities for the development of future office environments.Item Open Access Efficient fault tolerance for selected scientific computing algorithms on heterogeneous and approximate computer architectures(2018) Schöll, Alexander; Wunderlich, Hans-Joachim (Prof. Dr.)Scientific computing and simulation technology play an essential role to solve central challenges in science and engineering. The high computational power of heterogeneous computer architectures allows to accelerate applications in these domains, which are often dominated by compute-intensive mathematical tasks. Scientific, economic and political decision processes increasingly rely on such applications and therefore induce a strong demand to compute correct and trustworthy results. However, the continued semiconductor technology scaling increasingly imposes serious threats to the reliability and efficiency of upcoming devices. Different reliability threats can cause crashes or erroneous results without indication. Software-based fault tolerance techniques can protect algorithmic tasks by adding appropriate operations to detect and correct errors at runtime. Major challenges are induced by the runtime overhead of such operations and by rounding errors in floating-point arithmetic that can cause false positives. The end of Dennard scaling induces central challenges to further increase the compute efficiency between semiconductor technology generations. Approximate computing exploits the inherent error resilience of different applications to achieve efficiency gains with respect to, for instance, power, energy, and execution times. However, scientific applications often induce strict accuracy requirements which require careful utilization of approximation techniques. This thesis provides fault tolerance and approximate computing methods that enable the reliable and efficient execution of linear algebra operations and Conjugate Gradient solvers using heterogeneous and approximate computer architectures. The presented fault tolerance techniques detect and correct errors at runtime with low runtime overhead and high error coverage. At the same time, these fault tolerance techniques are exploited to enable the execution of the Conjugate Gradient solvers on approximate hardware by monitoring the underlying error resilience while adjusting the approximation error accordingly. Besides, parameter evaluation and estimation methods are presented that determine the computational efficiency of application executions on approximate hardware. An extensive experimental evaluation shows the efficiency and efficacy of the presented methods with respect to the runtime overhead to detect and correct errors, the error coverage as well as the achieved energy reduction in executing the Conjugate Gradient solvers on approximate hardware.Item Open Access Eine Methode zum Verteilen, Adaptieren und Deployment partnerübergreifender Anwendungen(2022) Wild, Karoline; Leymann, Frank (Prof. Dr. Dr. h. c.)Ein wesentlicher Aspekt einer effektiven Kollaboration innerhalb von Organisationen, aber vor allem organisationsübergreifend, ist die Integration und Automatisierung der Prozesse. Dazu zählt auch die Bereitstellung von Anwendungssystemen, deren Komponenten von unterschiedlichen Partnern, das heißt Abteilungen oder Unternehmen, bereitgestellt und verwaltet werden. Die dadurch entstehende verteilte, dezentral verwaltete Umgebung bedarf neuer Konzepte zur Bereitstellung. Die Autonomie der Partner und die Verteilung der Komponenten führen dabei zu neuen Herausforderungen. Zum einen müssen partnerübergreifende Kommunikationsbeziehungen realisiert und zum anderen muss das automatisierte dezentrale Deployment ermöglicht werden. Eine Vielzahl von Technologien wurde in den letzten Jahren entwickelt, die alle Schritte von der Modellierung bis zur Bereitstellung und dem Management zur Laufzeit einer Anwendung abdecken. Diese Technologien basieren jedoch auf einer zentralisierten Koordination des Deployments, wodurch die Autonomie der Partner eingeschränkt ist. Auch fehlen Konzepte zur Identifikation von Problemen, die aus der Verteilung von Anwendungskomponenten resultieren und die Funktionsfähigkeit der Anwendung einschränken. Dies betrifft speziell die partnerübergreifenden Kommunikationsbeziehungen. Um diese Herausforderungen zu lösen, stellt diese Arbeit die DivA-Methode zum Verteilen, Adaptieren und Deployment partnerübergreifender Anwendungen vor. Die Methode vereinigt die globalen und lokalen Partneraktivitäten, die zur Bereitstellung partnerübergreifender Anwendungen benötigt werden. Dabei setzt die Methode auf dem deklarativen Essential Deployment Meta Model (EDMM) auf und ermöglicht damit die Einführung deploymenttechnologieunabhängiger Modellierungskonzepte zur Verteilung von Anwendungskomponenten sowie zur Modellanalyse und -adaption. Das Split-and-Match-Verfahren wird für die Verteilung von Anwendungskomponenten basierend auf festgelegten Zielumgebungen und zur Selektion kompatibler Cloud-Dienste vorgestellt. Für die Ausführung des Deployments können EDMM-Modelle in unterschiedliche Technologien transformiert werden. Um die Bereitstellung komplett dezentral durchzuführen, werden deklarative und imperative Technologien kombiniert und basierend auf den deklarativen EDMM-Modellen Workflows generiert, die die Aktivitäten zur Bereitstellung und zum Datenaustausch mit anderen Partnern zur Realisierung partnerübergreifender Kommunikationsbeziehungen orchestrieren. Diese Workflows formen implizit eine Deployment-Choreographie. Für die Modellanalyse und -adaption wird als Kern dieser Arbeit ein zweistufiges musterbasiertes Verfahren zur Problemerkennung und Modelladaption eingeführt. Dafür werden aus den textuellen Musterbeschreibungen die Problem- und Kontextdefinition analysiert und formalisiert, um die automatisierte Identifikation von Problemen in EDMM-Modellen zu ermöglichen. Besonderer Fokus liegt dabei auf Problemen, die durch die Verteilung der Komponenten entstehen und die Realisierung von Kommunikationsbeziehungen verhindern. Das gleiche Verfahren wird auch für die Selektion geeigneter konkreter Lösungsimplementierungen zur Behebung der Probleme angewendet. Zusätzlich wird ein Ansatz zur Selektion von Kommunikationstreibern abhängig von der verwendeten Integrations-Middleware vorgestellt, wodurch die Portabilität von Anwendungskomponenten verbessert werden kann. Die in dieser Arbeit vorgestellten Konzepte werden durch das DivA-Werkzeug automatisiert. Zur Validierung wird das Werkzeug prototypisch implementiert und in bestehende Systeme zur Modellierung und Ausführung des Deployments von Anwendungssystemen integriert.Item Open Access Visual analysis of sequential data(2025) Munz-Körner, Tanja; Weiskopf, Daniel (Prof. Dr.)Sequential and temporal data is omnipresent in various areas of our lives. It is characterized by a sequence of data points in a fixed order, possibly with a temporal component. With an increasing amount of data being generated and collected, and different types of data originating from various domains, appropriate methods are needed to examine, interpret, understand, and draw conclusions from complex processes. Depending on the use case, the amount of data, and the target group, different analysis methods have to be chosen or developed. While visualization alone can already provide interesting insights into the data, interactive visual analysis helps users extract additional information by letting them focus on specific parts of the data and exploring it from different perspectives. Techniques such as brushing and linking and multiple coordinated views (multiple visualizations for the same data that are linked) help realize such an examination. In this thesis, several approaches for visually analyzing sequential data are presented. The focus lies particularly on two key application areas: eye tracking and the interpretability of machine learning (ML) methods. Additionally, the use of dimensionality reduction methods during preprocessing for visualization is an important concept of this work. In all these areas, sequential or temporal components play important roles. They can be the subject of exploration, used as input data to trigger complex processes, represent internal mechanisms within methods, or be the output of a process. Users may want to examine or compare them to understand the data better. In the area of eye tracking analysis, this thesis presents a visual analysis approach that addresses the influence of various filter settings (parameter choices) on the data being visualized and interpreted. Additionally, a method is presented that combines temporal data from different sources to enable a better comparison of this data. Preprocessing steps play a crucial role in both methods to allow meaningful visualizations of the data and subsequent examination of the data. Next, various ML approaches are considered. The interpretability of ML techniques is currently a very important and challenging topic. Especially ML models in the area of natural language processing (NLP) deal with sequential components as input data, and also, the internal operations follow sequential processing steps. This thesis demonstrates that, in the field of NLP, internal information from neural machine translation (NMT), visual question answering (VQA), and text classification tasks can be made available to users for an enhanced understanding of internal mechanisms and to improve prediction results. Toward the end of this thesis, dimensionality reduction techniques are applied as a preparation step for visualizing sequential data. First, dimensionality reduction is used in an interactive system to examine text classification in the context of ML. However, interpreting 2D visualizations of dimensionally reduced sequential data requires careful consideration due to the possibility of data loss, misleading projections, and potential misinterpretation of the visualization itself. Therefore, in this work, visualization approaches are presented that address this challenge to provide methods to prevent misinterpretation. Overall, all presented interactive visualization approaches of this thesis use sequential data as input, and the visual analysis techniques help users during data exploration, interpretation, for debugging purposes, or to improve prediction results generated with ML models.Item Open Access Elastic parallel systems for high performance cloud computing(2020) Kehrer, Stefan; Blochinger, Wolfgang (Prof. Dr.)High Performance Computing (HPC) enables significant progress in both science and industry. Whereas traditionally parallel applications have been developed to address the grand challenges in science, as of today, they are also heavily used to speed up the time-to-result in the context of product design, production planning, financial risk management, medical diagnosis, as well as research and development efforts. However, purchasing and operating HPC clusters to run these applications requires huge capital expenditures as well as operational knowledge and thus is reserved to large organizations that benefit from economies of scale. More recently, the cloud evolved into an alternative execution environment for parallel applications, which comes with novel characteristics such as on-demand access to compute resources, pay-per-use, and elasticity. Whereas the cloud has been mainly used to operate interactive multi-tier applications, HPC users are also interested in the benefits offered. These include full control of the resource configuration based on virtualization, fast setup times by using on-demand accessible compute resources, and eliminated upfront capital expenditures due to the pay-per-use billing model. Additionally, elasticity allows compute resources to be provisioned and decommissioned at runtime, which allows fine-grained control of an application's performance in terms of its execution time and efficiency as well as the related monetary costs of the computation. Whereas HPC-optimized cloud environments have been introduced by cloud providers such as Amazon Web Services (AWS) and Microsoft Azure, existing parallel architectures are not designed to make use of elasticity. This thesis addresses several challenges in the emergent field of High Performance Cloud Computing. In particular, the presented contributions focus on the novel opportunities and challenges related to elasticity. First, the principles of elastic parallel systems as well as related design considerations are discussed in detail. On this basis, two exemplary elastic parallel system architectures are presented, each of which includes (1) an elasticity controller that controls the number of processing units based on user-defined goals, (2) a cloud-aware parallel execution model that handles coordination and synchronization requirements in an automated manner, and (3) a programming abstraction to ease the implementation of elastic parallel applications. To automate application delivery and deployment, novel approaches are presented that generate the required deployment artifacts from developer-provided source code in an automated manner while considering application-specific non-functional requirements. Throughout this thesis, a broad spectrum of design decisions related to the construction of elastic parallel system architectures is discussed, including proactive and reactive elasticity control mechanisms as well as cloud-based parallel processing with virtual machines (Infrastructure as a Service) and functions (Function as a Service). To evaluate these contributions, extensive experimental evaluations are presented.Item Open Access Generative models and domain adaptation for autonomous driving(2024) Eskandar, George; Yang, Bin (Prof. Dr.-Ing.)Artificial Intelligence (AI) and Deep Learning (DL) have recently affected human society in profound ways, sparking conversations about their technological, social and ethical impacts on our daily lives. The development of intelligent agents capable of perceiving, reasoning, and interacting with the 3D spaces is crucial, especially for Autonomous Driving (AD), which promises to revolutionize mobility, reduce accidents, and conserve time and energy. However, achieving full AD is hindered by the challenge of generalizing to new conditions. This is because autonomous vehicles rely on DL models which are limited by the scope of their training data. The sheer variety of potential real-world driving situations, particularly dangerous ones, cannot be reproduced for training purposes. When encountering these unrepresented situations, the vehicles face a domain gap, where they must operate in conditions different from what they were trained on. This mismatch can undermine their safety and dependability, restricting their practical use and leading to significant financial setbacks for car manufacturers. Research efforts against domain gaps have been channeled into two main directions: (1) employing generative AI models to produce synthetic data, thus augmenting the training datasets, and (2) fine-tuning pre-trained DL models for data in new domains without the need for manual labeling. The former strategy is known as generative models, while the latter is referred to as domain adaptation. However, current approaches suffer from multiple drawbacks when applied to AD in particular. For instance, generative models struggle to achieve photorealism, controllability and label-efficiency at the same time, when applied to complex scenes. On the other hand, domain adaptation is still understudied for some sensor modalities like LiDAR and for sensor fusion models (camera and LiDAR) which are widely used in AD, limiting their potential. This dissertation is part of the KI Delta Learning project, funded by the Bundesministerium für Wirtschaft und Energie (BMWi), to address the critical challenge of domain gaps in AD. Towards this goal, we developed novel approaches in three key AD areas: (1) Generating photorealistic and editable urban scenes, (2) enhancing the resolution of LiDAR pointclouds and (3) adapting 2D and 3D object detectors to new domains. In the first two applications, we developed novel generative models that provide additional training data (camera and LiDAR). In the third area, we established new architectures and training strategies to build models that are more robust against domain shifts. Across all areas, the considered domain gaps encompass weather, sensor and location changes. In our first application, we devised a series of models capable of producing high-quality, photorealistic images from semantic maps, tailored to different annotation cost levels. For the lowest cost, we introduced two fully unsupervised models: Unsupervised Semantic Image Synthesis (USIS) and Synthetic-to-Real SIS. USIS operates on unpaired images and semantic maps, ideally where both share comparable spatial and semantic characteristics derived from real-world data. The Synthetic-to-Real SIS model mitigates the need for such similarity by accommodating labels generated through computer graphics, which may differ statistically from real-world imagery. We then developed a semi-supervised model, Semi-Paired SIS, which learns from a vast collection of unpaired images and labels, plus a smaller subset of paired data. Semi-Paired SIS nearly matches the performance of fully supervised approaches with significantly less paired data. Lastly, we introduced a supervised model, Urban-StyleGAN, capable of generating images and labels from noise vectors and modifying the image through vector manipulation. In the second application, we developed a novel model to upsample low-resolution LiDAR pointclouds into high-resolution, balancing cost-effectiveness and performance. In the third application, we pioneered a model to adapt a multi-sensor 2D object detector to harsh weather conditions. Finally, a large empirical study on the robustness of 3D object detectors was conducted, yielding several important novel findings in the robustness and adaptation to unseen conditions. Each developed model was rigorously tested across multiple public benchmarks, consistently achieving state-of-the-art results. In conclusion, this dissertation presents significant theoretical and practical advancements in generative models and domain adaptation for AD. The important benefits of this work encompass enhanced photorealism, improved controllability, greater label efficiency, and increased robustness against domain shifts, all of which contribute to the safety and reliability of autonomous systems. We hope our contributions can benefit the DL and AD communities and find applications in other related fields (medical, satellite image processing, radar signal processing...), fostering innovation and practical advancements across these fields.