Digital Library of Gesellschaft für Informatik e.V.
Not a member yet
40758 research outputs found
Sort by
Virtual Reality in Speech Therapy Education – A new approach to teach the Flexible Endoscopic Evaluation of Swallowing (FEES)
No full textRecent studies have demonstrated the efficacy of virtual reality as a powerful tool for clinical training in healthcare. It offers a safe and controlled environment where students can develop essential hands-on skills without exposing patients to risk. In the domain of speech therapy, the utilisation of this technology for educational purposes remains limited. Virtual reality holds considerable potential in the training of students in the performance of complex procedures, such as the Flexible Endoscopic Evaluation of Swallowing (FEES), which require a high level of precision, familiarity with instrumentation, and an efficient clinical decision-making process. Our research project aims at developing a virtual reality training module able to empower students’ knowledge, foster their decision-making process and enhance their clinical skills while gaining hands-on experience
Sonic Map Explorer – Ein emergentes Interface zur Live-Interaktion mit KI-Musiksystemen
No full textSonic Map Explorer ist ein emergentes Interface zur intuitiven Live-Interaktion mit generativen KI-Musiksystemen. Es projiziert Audiosignale in Echtzeit auf eine zweidimensionale Klangkarte, die gleichzeitig visuelles Feedback und Steuerdaten für generative KI bietet. Durch ein kombiniertes Verfahren aus spektraler Audioanalyse, UMAP-Dimensionsreduktion und neuronaler Mapping-Architektur verzichtet das System auf vordefinierte Parameter oder Prompts und ermöglicht stattdessen eine Klanggesten-gesteuerte Co-Kreation. Das System versteht sich als prototypischer Beitrag zur kollaborativen Mensch-KI-Interaktion in der musikalischen Improvisation
A Jupyter Book Template for Research-Based Open Educational Resources in Data Literacy
No full textThis paper introduces a Jupyter Book–based template designed to support the creation of modular, interactive Open Educational Resources (OER) for data literacy. The template provides educators with a structured framework to convert authentic research workflows into pedagogically coherent modules, each comprising theoretical background, hands-on practice, reflective activities, and formative assessment. By embedding data skills within discipline-specific case studies, it enables the development of reusable and accessible learning materials. We describe the template’s instructional design and technical architecture, and highlight the core challenge of aligning software requirements with the diverse software competencies of OER creators
Understanding Knowledge Management Systems: Psychological Needs As A Conceptual Lens On User Experience, Research, and Design
No full textKnowledge Management Systems (KMS) are crucial for capturing,
sharing, and applying organizational knowledge. While technical
and organizational aspects of KMS have been discussed in the HCI
literature, the psychological dimension has received less attention.
Still psychological need fulfillment and adequately addressing the
motivations of different stakeholders could be a deciding factor for
their actual usage and success. This paper introduces psychological
needs (e.g. autonomy, competence, relatedness, and meaning) as a
valuable framework to better select, evaluate, and design KMS. First,
the complex landscape of KMS types, use cases and stakeholders
is mapped, demonstrating that these classifications often fall short
of practical realities. Then psychological needs are proposed as
an integrative lens to make sense of this complexity and guide
both research and design decisions. This perspective is particularly
relevant for UX researchers, designers, and organizational leaders
or users in general seeking to foster more suitable Knowledge
Management Systems
A Neural Network Framework for Plasticity and Damage Evolution in Concrete
No full textAccurately capturing the nonlinear material behavior like damage evolution and cyclic plasticity remains a central challenge in computational mechanics [AWK21]. Traditional finite element approaches [XG16] often exhibit mesh sensitivity and require uniaxial tensile testing data, which requires an iterative and indirect process to determine suitable model parameters for varying loading conditions. Previous studies have shown the potential of neural material modeling and data-driven modeling for complex materials [Hi25; HK24]. Accordingly, to allow for a simplified material modeling process, a neural network (NN)-based framework is proposed for the material behavior of concrete using a history-dependent NN. The case study presented for validation here includes surrogate data obtained by numerical experiments with a finite element (FEM) software using the Concrete Damaged Plasticity (CDP) model and covering tensile, compressive, and cyclic loading cases. A feed-forward neural network is trained to predict the incremental plastic strain and damage parameters based on a recurrent information flow. The results demonstrate that the NN generalizes well across both training and testing load cases and offers a promising alternative to conventional material models. The network captures the nonlinear plastic strain and damage evolution with high accuracy and computational efficiency
Mosul: Preserving the After-Effects of War
No full textThis project preserves the after-effects of war in Mosul by utilizing Virtual Reality (VR) and Augmented Reality (AR) technologies. A virtual environment has been created to display the current state of Mosul’s old area, which suffered the most damage due to war. Additionally, we integrate comparison images to provide users with a visual contrast between the past and present states of these locations. Furthermore, we have developed a virtual tour that allows users to explore these reconstructed sites interactively, enhancing their understanding of Mosul’s historical and cultural significance
Automatic Ergonomics Evaluation of Sawing and Hammering Activities Using a Wrist-Worn IMU
No full textTens of millions of workers around the world suffer from occupational diseases related to Repetitive Strain Injuries (RSI), caused by non-ergonomic manual labor. This paper presents an application for automatically evaluating the ergonomics of repetitive hammering and sawing activities using a wrist-worn IMU. The following application automatically detects the time spent and the number of strokes per shift for hammering and sawing activities, and outputs the Occupational Repetitive Action (OCRA) index as per ISO 11228-3:2007 standard on handling low loads at high frequency. The proposed framework involves using Machine Learning models and biomechanical constraints to automatically calculate the OCRA index, which, if exceeds a certain threshold, indicates that the worker has to make some ergonomic changes to their routine starting from the next shift, to avoid chronic health risks. Authors demonstrate the feasibility and robustness of this application by employing the system in an experimental study
Lernendenzentriertes Onboarding in RSE: Ein Forschungsdesign zur Persona-Entwicklung
No full textDieser Posterbeitrag stellt ein geplantes Forschungsdesign vor, das Lernende mit heterogenem Vorwissen in den Mittelpunkt stellt. Ziel ist es, mittels eines strukturierten Fragebogens Informationen zu Motivation, Berufszielen, Herausforderungen und Weiterbildungsbedarfen von Personen im Research Software Engineering (RSE)-Kontext zu erheben. Diese Erkenntnisse sollen zur Entwicklung von Personas beitragen, die als Grundlage für zielgruppenspezifisches Onboarding-Material und didaktische Konzepte dienen können. Der Beitrag regt eine Diskussion über Lernende im RSE-Bereich an – eine Gruppe, die häufig interdisziplinär geprägt ist und vor besonderen Herausforderungen steht. Anders als klassische Softwareentwickler*innen verfügen RSE-Lernende nicht immer über formale Ausbildung in Software Engineering, müssen aber dennoch robuste, nachhaltige und vertrauenswürdige Forschungssoftware entwickeln. Dies erfordert Weiterbildungsangebote, die flexibel, niedrigschwellig und auf Teilhabe ausgerichtet sind. Da viele dieser Lernenden zeitlich eingebunden sind, kommt Konzepten des selbstgesteuerten eLearnings mit hoher Verfügbarkeit und modularem Aufbau besondere Bedeutung zu. Mit dem vorgestellten Fragebogen sollen auch Einflussfaktoren auf ein gelungenes Onboarding identifiziert werden – im Sinne einer praxisnahen, nutzendenzentrierten Weiterbildung, die der Vielfalt in der RSE-Community gerecht wird
Face Off: External Tracking vs. Manual Control for Facial Expressions in Multi-User Extended Reality
No full textIn distributed multi-user XR spaces, avatar facial expressions are usually enabled by built-in sensors in high-end HMDs. Motivated by the diverse landscape of devices without these capabilities, we investigate two alternative methods to execute facial expressions. In a study with 18 participants collaborating in dyads, we compared (1) external webcam-based face tracking and (2) manually triggered preset expressions, exploring trade-offs between less reliable, video-based tracking of partially obscured faces and less natural manual control. Our results show that participants prefer manual triggering over unstable face tracking, as the latter leads to significantly higher task load and effort, while the former did not negatively influence interpersonal communication and was easier to use