Open Access Zeitschriften des Universitätsverlages der Technischen Universität Chemnitz
Not a member yet
466 research outputs found
Sort by
Evaluation of Propeller Inspection Using Different Deployment Strategies
Get PDFIn recent years, the use of Unmanned Aerial Vehicles (UAVs) for various applications has increased significantly. Among these applications, the inspection of infrastructures using UAVs has become a prominent area of research. This paper evaluates the efficiency of the YOLOv5 algorithm for propeller inspection. The algorithm\u27s deployment across various platforms such as PC, Google Colab, and Jetson Nano is examined, with a focus on different deployment formats like PyTorch, ONNX, TensorFlow Lite, and others. The studyhighlights the often-overlooked importance of the deployment phase in the development of AI models and underscores its significance for the practical application of AI in real-world scenarios.Keywords— Computer vision, algorithm deployment, propeller inspection, Deployment strategies, efficiency improvemen
Diagramme von Gesten. Eine zeichentheoretische Analyse digitaler Bewegungsspuren
Get PDFZusammenfassung. Dieser Beitrag versucht eine zeichentheoretische Untersuchung von Diagrammen, wie sie im Rahmen der motion-capture-gestützten, empirischen Gestenforschung zur Repräsentation und Analyse redebegleitender Gesten erzeugt und verwendet werden. Nach Vorüberlegungen zum Diagrammbegriff und einer Darlegung herkömmlicher Mittel der visuellen Repräsentation von redebegleitender Gestik, folgt eine zeichentheoretische Analyse des Motion-Capture-Verfahrens und deren diagrammatisch-semiotische Interpretation. Motion-Capture-Diagramme sind, so unsere Hypothese, das Produkt eines diagrammatischen Verfahrens, welches als empirische Skala zu deuten ist. Die „Erkenntniskraft der Linie“ (Krämer 2011) liegt hier in der ikonischen Visualisierung von Trajektorien, welche durch die Projektion eines Indexes auf ein Koordinatensystem zu Perzeptionsgestalten von Gesten werden. Weiterhin erörtert dieser Beitrag diagrammatische Prozesse in Gesten selbst, welche das eingangs diskutierte System als holistische Gestalt sichtbar machen kann. Diese Verfahren finden ihren räumlich-aktionalen Niederschlag u.a. sowohl in verkörperten Bildschemata, in der relationalen Repräsentation von Abstrakta, als auch in Gestendiagrammen basierend auf mentalen Karten, wie sie z.B. bei der Planung von Reisen in der Face-to-Face-Kommunikation im Gestenraum entstehen.Summary. This paper presents a semiotic investigation of how diagrams are produced and used in empirical, motion-capture based gesture research for the purposes of representation and analysis of co-speech gestures. First, we discuss the notion of diagram, drawing on Peircean semiotics and more recent accounts of diagrammatic thinking, and traditional ways of recording and representing co-speech gestures. We then employ analytic tools stemming from semiotic theory to a) account for the notational procedures used in motion-capture technology and b) offer a diagrammatic interpretation of the signs resulting from them. Our hypothesis is that motion-capture diagrams are products of operative diagrammatic processes and that they are comparable with empirical measurements. The “epistemic force of the trace” (Krämer 2011) consists in the iconic visualization of movement trajectories which become perceptual gestural gestalts through projecting an indexical sign onto a coordinate system that virtually structures gesture space. We further highlight diagrammatic processes underpinning co-speech gestures themselves, which may be visualized as holistic gestalts with the help of optical motion-capture technology. These cognitive-semiotic processes manifest themselves dynamically and spatially through embodied image schemas, relational representations of abstract concepts, and gestural diagrams based on mental maps (for example, when conversational partners are jointly planning a journey)
Exaktheit – Eindeutigkeit – Eigentlichkeit. Zur semiotischen Explikation terminologischer Grundeigenschaften
Get PDFZusammenfassung. Die terminologischen Eigenschaften Exaktheit, Eindeutigkeit und Eigentlichkeit können anhand des semiotischen Dreiecks von Ogden und Richards erfasst und aufeinander bezogen werden: Mit Blick auf Bedeutung, Ausdruck und Wirklichkeit erweisen sie sich dabei als referentielle, semantische und mentale (motivierte) Unmissverständlichkeit von Termini. Sie werden im Rahmen eines systemlinguistischen Inventarmodells, eines pragmalinguistischen Kontextmodells und eines kognitionslinguistischen Funktionsmodells jeweils unterschiedlich konzipiert. Exaktheit, Eindeutigkeit und Eigentlichkeit werden systemlinguistisch postuliert, pragmalinguistisch analysiert oder kognitionslinguistisch interpretiert. Argumentativ wie historisch zeigt sich hierbei ein dialektisches Verhältnis der drei Modelle.
Summary. The terminological properties of exactness, uniqueness and authenticity can be related to each other on the basis of Ogden and Richard’s semiotic triangle. With respect to meaning, expression, and reality they have to be understood as referential, semantic, and motivated comprehensibility of terms. They are conceptualized in different ways within the framework of an inventory model in the tradition of system-oriented linguistics, a context model in the tradition of pragmalinguistics, and a functional model in the tradition of cognitive linguistics. Exactness, uniqueness and authenticity are systematically postulated, pragmatically analyzed, or cognitively interpreted. Argumentatively and historically, the approach demonstrates a dialectical relationship between the three models
Learning IoT Course Using Web Platform Technology
Get PDFThe IoT market of Kazakhstan is, to some extent, the heir to technological solutions that have been actively developing in the republic for more than a decade. Improvements in devices allow them to hear, see, think, and sometimes act. Today, with the development of this technology, special courses in this area have begun to be introduced. Thus, the traditional model of university education is changing. The article pre- sents the technology for organizing project-based learning IoT course on the basis of the Faculty of Information Technology. L.N. Gumilyov at the Eurasian National University. Based on the analysis of scientific, methodo-logical and regulatory articles, a structural model for organizing IoT course training through web technologies. In the course of studying the IoT course from various educational platforms, students encountered various problems and found solutions to them. As a result, we have created an educational platform for learning IoT. In the educational platform, the participants learned to apply theoretical knowledge to solve real problems in practice and gained an unforgettable team experience
Developing Information Competences of the Students in Technical Direction with Helping the Technology of “Network Boomerang” Principles
Get PDFThe article discusses ways of increasing students\u27 engagement, motivation and participation in higher education using "Network Boomerang" technology, as an example of pre-service IT teacher training. With the onset of the global pandemic, the role of digital technologies has significantly increased, highlighting the importance of digital transformation, particularly for developing countries, including Uzbekistan. Indeed, digital transformation has the potential to further modernize society and integrate the national economy into global processes. In this context, within the framework of ongoing reforms and the Development Strategy of the New Uzbekistan for the next five years, special attention is being given to the digitization of key areas of activity and the establishment of a genuine information society in the country
Zeichen im öffentlichen Raum: Funktionalisierung, Ästhetisierung und Mediatisierung
Get PDFDas Heft beschäftigt sich mit semiotischen Aspekten der Metropolenforschung. Urbane Zeichen reichen von Verkehrszeichen und Straßennamen über Denkmäler, Street Art und Werbung bis hin zu semiotischen Aspekten von Architektur und Stadtgestaltung. Die Beiträge des Hefts beschäftigen sich u.a. mit der Erinnerungskultur von Denkmälern und Straßennamen, der sich verändernden ästhetischen Wahrnehmung von Leuchtreklamen, politischer Protestkommunikation, phonetischen Transkriptionen in Graffiti und Werbung sowie QR-Codes als Verknüpfung von städtischen und digitalen bzw. virtuellen Räumen.This issue deals with semiotic aspects of metropolitan studies. Urban signs range from traffic signs and street names to monuments, street art and advertising to semiotic aspects of architecture and urban design. The issue\u27s contributions deal with the memory culture of monuments and street names, the changing aesthetic perception of neon signs, political protest communication, phonetic transcriptions in graffiti and advertising, and QR codes as a link between urban and digital or virtual spaces
Damping-induced dispersion in simple waveguides
No full textThe goal is to gain a non-dimensional formulation of the complex wavenumber knowing the onedimensional partial differential equation for wave propagation and the harmonic wave approach. This can be disassembled into its components, namely the imaginary part, showing the decay of a wave, and the real part, showing its spatial propagation. Having an eye on damping in waveguides including internal and external damping, we discuss the resulting consequences, which are frequency-dependent phase and group velocities (dispersion), as well as the energy transport and dissipation. This bears relevance for many domains of physics such as seismic and electromagnetic waves
Driving Behaviors Recognition Using Deep Neural Networks
Get PDFRoad accidents are skyrocketing, and traffic safety is a severe problem around the world. Many road traffic deaths are related to drivers’ unsafe behaviors. In this paper, we propose two different deep-learning models which classify the driver’s actions in a 60-second time frame into two main categories: Normal and Aggressive driving based on GPS data collected at 1 Hz, which is later preprocessed and passed to the proposed models to identify dominant driving behavior in each time frame. The models achieved an accuracy of 93.75 percent in real-world tests, which proves the efficiency of this method in driving behavior recognition