Cornucopia: tool support for selecting machine learning lifecycle artifact management systems. - Setúbal : Scitepress. - 1 Online-Ressource (Seite 444-450)Online-Ausgabe: Proceedings of the 18th International Conference on Web Information Systems and Technologies, WEBIST, October 25-27, 2022, in Valletta, Malta / editors: Stefan Decker ... - Setúbal : Scitepress, 2022, ISBN 978-989-758-613-2
The explorative and iterative nature of developing and operating machine learning (ML) applications leads to a variety of ML artifacts, such as datasets, models, hyperparameters, metrics, software, and configurations. To enable comparability, traceability, and reproducibility of ML artifacts across the ML lifecycle steps and iterations, platforms, frameworks, and tools have been developed to support their collection, storage, and management. Selecting the best-suited ML artifact management systems (AMSs) for a particular use case is often challenging and time-consuming due to the plethora of AMSs, their different focus, and imprecise specifications of features and properties. Based on assessment criteria and their application to a representative selection of more than 60 AMSs, this paper introduces an interactive web tool that enables the convenient and time-efficient exploration and comparison of ML AMSs.
https://doi.org/10.5220/0011591700003318
Raw filtering of JSON data on FPGAs. - In: Proceedings of the 2022 Design, Automation & Test in Europe Conference & Exhibition (DATE 2022), (2022), S. 250-255
https://doi.org/10.23919/DATE54114.2022.9774696
Towards more effective identification keys: a study of people identifying plant species characters. - In: People and nature, ISSN 2575-8314, Bd. 4 (2022), 6, S. 1603-1615
Accurate species identification is essential for ecological monitoring and biodiversity conservation. Interactive plant identification keys have been considerably improved in recent years, mainly by providing iconic symbols, illustrations, or images for the users, as these keys are also commonly used by people with relatively little plant knowledge. Only a few studies have investigated how well morphological characteristics can be recognized and correctly identified by people, which is ultimately the basis of an identification key's success. This study consists of a systematic evaluation of people's abilities in identifying plant-specific morphological characters. We conducted an online survey where 484 participants were asked to identify 25 different plant character states on six images showing a plant from different perspectives. We found that survey participants correctly identified 79% of the plant characters, with botanical novices with little or no previous experience in plant identification performing slightly worse than experienced botanists. We also found that flower characters are more often correctly identified than leaf characteristics and that characters with more states resulted in higher identification errors. Additionally, the longer the time a participant needed for answering, the higher the probability of a wrong answer. Understanding what influences users' plant character identification abilities can improve the development of interactive identification keys, for example, by designing keys that adapt to novices as well as experts. Furthermore, our study can act as a blueprint for the empirical evaluation of identifications keys. Read the free Plain Language Summary for this article on the Journal blog.
https://doi.org/10.1002/pan3.10405
Methods and tools of video monitoring for remote laboratory. - In: Building bridges towards responsible smart cities, (2022), S. 184-189
The ability to monitor the progress of online experiments using a video monitoring system is one of the most important advantages of remote laboratories. In this case, special requirements are imposed on the capture area, quality and degree of image details. Therefore, the further development of methods and tools of capturing, broadcasting, receiving and displaying video images to improve the quality of educational services provided by a remote laboratory is an urgent task. In the course of the work, technologies for capturing and processing video images, implementing wireless protocols for receiving (commands, requests) and transmitting data (images), as well as architectural features, principles of software implementation and operation of video monitoring systems are investigated and realized. The developed system provides the ability to organize video monitoring for remote experiments observing and fixing. Thus, students receive the fullest information and data streaming provides the effect of physical presence at the site of the experiment. The hardware/software complex is developed on the basis of ESP-32 and ESP-32-CAM modules, as well as Wi-Fi, HTTP and WebSocket protocols. The ability to update the developed software is implemented using the OtaDrive cloud service. This makes it easier and faster to deploy and test new software. The main advantages of the proposed solution are ease of assembly, customization, availability of components, the possibility of easy modernization of both hardware and software parts of the complex, low power consumption. It has low cost, which ensures availability even for small educational laboratories. The practical value of the work lies in the fact that the created software/hardware complex can be used to organize effective video monitoring of experiments in remote laboratories, as well as for other video surveillance systems in various fields.
https://doi.org/10.1109/E-TEMS53558.2022.9944406
3D INS/UWB based real time sensor fusion indoor position tracking architecture. - In: 2022 IEEE 13th Annual Information Technology, Electronics and Mobile Communication Conference (IEMCON), (2022), S. 94-101
Accurate indoor position tracking and analysis of the movement dynamics of autonomous driving systems are important challenges when it comes to automatize industrial processing, supply chains or warehouses. In this paper, the authors present an indoor position tracking architecture with a novel sensor fusion approach for autonomous robots in three-dimensional space. For robots to be able to drive autonomous, they need information of their position, speed and orientation in 3D-space. With the presented architecture, position information is provided by the Indoor Positioning System (IPS) and orientation information as well as velocity are determined by the Inertial Navigation System (INS). The proposed tracking architecture combines those informations with a sensor fusion approach, thus enabling the autonomous driving system.
https://doi.org/10.1109/IEMCON56893.2022.9946462
Deep-learning API-feature specification tool for formative assessment in workshops. - In: FIE 2022, Uppsala, Sweden, (2022), insges. 9 S.
This Research to Practice Full Paper is driven by the question: In workshops to develop deep-learning API-features, how can trainer be effectively supported in specifying and formative assessing of individual tasks (i.e. the API-features) for each student? In formative assessment, the trainer’s main difficulty in specifying the tasks is to ensure that the challenges in a task are calibrated to the particular needs of a student at a particular time (i.e. the student’s available skills). While goals that are very high challenging lead to a student’s anxiety, goals that are too easy generate student’s disengagement and boredom. This paper presents a specification tool designed to support the trainers in practicing Hattie’s Visible Learning [1] pedagogy in dialogues for tasks specification and formative assessments. Its design enables two support functions in visualization; 1) a function to declare visually a task specification to make its mastery goals clear and progressively challenging, and 2) a function to refactor a task specification for visualizing formative assessment feedback in dialogues for reflection. Besides the two functions, its innovative design enables the flow model based analysis to understand student engagement. The analysis is according to Csikszentmihalyi’s flow model to diagnose, in dialogues for formative assessments, one of student's eight emotional states (i.e. boredom, apathy, worry, anxiety, control, arousal, and flow) in terms of challenge level and skill level. To verify the tool effectiveness in trainer-student Freirean dialogues, we conducted our experimental tests in workshops where students with different culture backgrounds, and different levels of prior skills have tasks to develop API-features based on deep-learning principles and algorithms. The algorithms selected in the test cases are for automatic classification of images or of text documents. The test results show the effective use of the tool to create conditions to overcome student’s negative emotions (e.g. worry or anxiety), and help him/her experience positive emotions (e.g. control or flow) in learning and mastering the algorithms; to put it differently, to create situations in initial specification and co-specification of tasks, where student’s skills are engaged by progressively higher challenges to understand the deep-learning based classifier algorithms.
https://doi.org/10.1109/FIE56618.2022.9962474
SparkCAD: caching anomalies detector for spark applications. - In: Proceedings of the VLDB Endowment, ISSN 2150-8097, Bd. 15 (2022), 12, S. 3694-3697
Developers of Apache Spark applications can accelerate their workloads by caching suitable intermediate results in memory and reusing them rather than recomputing them all over again every time they are needed. However, as scientific workflows are becoming more complex, application developers are becoming more prone to making wrong caching decisions, which we refer to as caching anomalies, that lead to poor performance. We present and give a demonstration of Spark Caching Anomalies Detector (SparkCAD), a developer decision support tool that visualizes the logical plan of Spark applications and detects caching anomalies.
https://doi.org/10.14778/3554821.3554877
What are we missing for effective remote laboratories?. - In: 2022 IEEE German Education Conference (GeCon), (2022), insges. 6 S.
Remote laboratories play an important role in modern education. Because of this, a multitude of different protocols and systems for building and integrating remote laboratories have been developed over the last years. This Paper analyses these different protocols and systems (SCORM / xAPI, IMS-LTI, IEEE 1876-2019, IVI / VISA, OPC-UA, Weblab-Deusto / LabsLand, GOLDi-Labs) and sorts them into different layers to build a protocol stack for implementing remote laboratories, accompanied by some examples of what is currently possible to implement. A gap in the protocol stack was identified between Controller and the whole Laboratory where no standardised protocol exists. Suggestions for the requirements of such a protocol are laid out. The new possibilities with such a protocol are presented by showing examples of new types of remote laboratories, followed by a short discussion of problems not yet solved by such a new protocol.
https://doi.org/10.1109/GeCon55699.2022.9942771
Adaptable digital labs - motivation and vision of the CrossLab project. - In: 2022 IEEE German Education Conference (GeCon), (2022), insges. 6 S.
The flexibility and performance of digital laboratory elements such as remote labs, VR/AR or simulations summarized under the term cross-reality labs (CrossLabs), can be seen with the development in last and has been proven under the pandemic situation. Even though the potential of cross-reality labs is obvious referring to availability and flexibility for the students, these didactic solutions remain isolated at universities as well as for individual users. The implementations are mostly so rigid that the individual didactic objectives are not interchangeable between different universities and disciplines, hence there is a lack of interoperability. The CrossLab project seeks to design didactical, technical, and organizational solutions for open digital lab objects linking student-centered teaching and a cross-university learning environment. Of importance thereby is the fact that teaching is not adaptable to the digital laboratory, but the laboratories are adaptable to the requirements of the teaching-learning setting. The four project partners are working on a cross-type and cross-element mixture of diverse types of laboratories for cross-disciplinary use in a cross-universities settings. Thereby, the project leans on existing digital laboratories in various disciplines to create an open teaching and learning environment which can be adapted to the needs of students and to provide students with the skills necessary for future working scenarios.
https://doi.org/10.1109/GeCon55699.2022.9942759
Elektrophysiologische Wirkung transokularer Ströme. - Ilmenau : Universitätsbibliothek, 2022. - 1 Online-Ressource (xxiv, 140 Blätter)
Technische Universität Ilmenau, Dissertation 2022
Die transkranielle Gleichstromstimulation mit schwachen Strömen (≤ 2 mA) ist ein etabliertes Verfahren zur Untersuchung des menschlichen Nervensystems und zur Behandlung psychologisch-neurologischer Erkrankungen. In den letzten Jahren hat die Stimulation des visuellen Systems, speziell der Retina, an Aufmerksamkeit gewonnen, da Hinweise auf positive Einflüsse einer transokularen Stromstimulation zur Behandlung neurodegenerativer retinaler Erkrankungen gefunden wurden. Bis heute fehlt jedoch ein Wirkungskonzept, das den Einfluss der Stromstimulation auf die verschiedenen retinalen Neuronen beschreibt. Aus Studien ist bekannt, dass ein visuell evoziertes Potential durch eine transkranielle Gleichstromstimulation des visuellen Cortex in den charakteristischen Amplituden beeinflusst werden kann. Daraus ergab sich das methodische Verfahren ein evoziertes Potential retinaler Herkunft, welches durch das Elektroretinogramm erfasst wird, in Kombination mit einer transokularen Stromstimulation zu nutzen, um den Einfluss der Stromstimulation auf die retinalen Zellen zu untersuchen. Durch Variation der visuellen Stimulationsparameter zur Generierung des Elektroretinogramms, ist es möglich, unterschiedliche retinale Neuronen in den Fokus zu nehmen. Im ersten Teil der Dissertation wurde ein grundlegendes Studienkonzept auf Basis des Verfahrens der experimentellen Prozessanalyse entwickelt. Aus diesem wurden Anforderungen an einen Mess- und Stimulationsplatz abgeleitet, um eine simultane transokulare Gleichstromstimulation und Aufnahme eines Elektroretinogramms zu ermöglichen. Des Weiteren wurden Anforderungen an das Studiendesign abgeleitet. Im zweiten Teil der Dissertation wurden insgesamt vier aufeinander aufbauende Probandenstudien durchgeführt, analysiert und interpretiert. In diesen wurden sowohl unterschiedliche retinale Neuronen als auch der Einfluss der Position der Stromstimulationselektroden untersucht. Aus den Ergebnissen wurde ein Wirkungskonzept für die transokulare Gleichstromstimulation auf die retinalen Neuronen abgeleitet. Dieses besagt, dass primär die retinalen Ganglienzellen während einer transokularen Gleichstromstimulation in ihrer elektrophysiologischen Aktivität beeinflusst werden. Die Ergebnisse der Dissertation lassen sich weiterführend für die Entwicklung und die Optimierung von Therapieverfahren zur Behandlung neurodegenerativer retinaler Erkrankungen sowie zur Entwicklung objektiver Messmethoden zum Nachweis einer transokularen Stromwirkung verwenden.
https://doi.org/10.22032/dbt.53716