Fachbereich Elektrotechnik und Informationstechnik
Refine
Year of publication
Institute
- ECSM European Center for Sustainable Mobility (25) (remove)
Has Fulltext
- no (25)
Document Type
- Conference Proceeding (16)
- Article (4)
- Part of a Book (3)
- Doctoral Thesis (1)
- Report (1)
Keywords
- Smart Building (2)
- 3D object detection (1)
- Android (1)
- Anomalieerkennung (1)
- BIM (1)
- Building Automation (1)
- CCD-Bildwandler (1)
- CO2 (1)
- Carbon Dioxide (1)
- Conpot (1)
This paper addresses the pixel based recognition of 3D objects with bidirectional associative memories. Computational power and memory requirements for this approach are identified and compared to the performance of current computer architectures by benchmarking different processors. It is shown, that the performance of special purpose hardware, like neurocomputers, is between one and two orders of magnitude higher than the performance of mainstream hardware. On the other hand, the calculation of small neural networks is performed more efficiently on mainstream processors. Based on these results a novel concept is developed, which is tailored for the efficient calculation of bidirectional associative memories. The computational efficiency is further enhanced by the application of algorithms and storage techniques which are matched to characteristics of the application at hand.
This paper addresses the pixel based classification of three dimensional objects from arbitrary views. To perform this task a coding strategy, inspired by the biological model of human vision, for pixel data is described. The coding strategy ensures that the input data is invariant against shift, scale and rotation of the object in the input domain. The image data is used as input to a class of self organizing neural networks, the Kohonen-maps or self-organizing feature maps (SOFM). To verify this approach two test sets have been generated: the first set, consisting of artificially generated images, is used to examine the classification properties of the SOFMs; the second test set examines the clustering capabilities of the SOFM when real world image data is applied to the network after it has been preprocessed to be invariant against shift, scale and rotation. It is shown that the clustering capability of the SOFM is strongly dependant on the invariance coding of the images.
Aim of the AXON2 project (Adaptive Expert System for Object Recogniton using Neuml Networks) is the development of an object recognition system (ORS) capable of recognizing isolated 3d objects from arbitrary views. Commonly, classification is based on a single feature extracted from the original image. Here we present an architecture adapted from the Mixtures of Eaqerts algorithm which uses multiple neuml networks to integmte different features. During tmining each neural network specializes in a subset of objects or object views appropriate to the properties of the corresponding feature space. In recognition mode the system dynamically chooses the most relevant features and combines them with maximum eficiency. The remaining less relevant features arz not computed and do therefore not decelerate the-recognition process. Thus, the algorithm is well suited for ml-time applications.
This paper describes the realization of a novel neurocomputer which is based on the concepts of a coprocessor. In contrast to existing neurocomputers the main interest was the realization of a scalable, flexible system, which is capable of computing neural networks of arbitrary topology and scale, with full independence of special hardware from the software's point of view. On the other hand, computational power should be added, whenever needed and flexibly adapted to the requirements of the application. Hardware independence is achieved by a run time system which is capable of using all available computing power, including multiple host CPUs and an arbitrary number of neural coprocessors autonomously. The realization of arbitrary neural topologies is provided through the implementation of the elementary operations which can be found in most neural topologies.
In der Vergangenheit basierten große Systemintegrationsprojekte in der Regel auf Individualentwicklungen für einzelne Kunden. Getrieben durch Kostendruck steigt aber der Bedarf nach standardisierten Lösungen, die gleichzeitig die individuellen Anforderungen des jeweiligen Umfelds berücksichtigen. T-Systems GEI GmbH wird beiden Anforderungen mit Produktkerneln gerecht. Neben den technischen Aspekten der Kernelentwicklung spielen besonders organisatorische Aspekte eine Rolle, um Kernel effizient und qualitativ hochwertig zu entwickeln, ohne deren Funktionalitäten ins Uferlose wachsen zu lassen. Umgesetzt hat T-Systems dieses Konzept für Flughafeninformationssysteme. Damit kann dem wachsenden Bedarf der Flughafenbetreiber nach einer effizienten und kostengünstigen Softwarelösung zur Unterstützung Ihrer Geschäftsprozesse entsprochen werden.
Der Erfolg eines Softwarenentwicklungsprojektes insbesondere eines Systemintegrationsprojektes wird mit der Erfüllung des „Teufelsdreiecks“, „In-Time“, „In-Budget“, „In-Quality“ gemessen. Hierzu ist die Kenntnis der Software- und Prozessqualität essenziell, um die Einhaltung der Qualitätskriterien festzustellen, aber auch, um eine Vorhersage hinsichtlich Termin- und Budgettreue zu treffen. Zu diesem Zweck wurde in der T-Systems Systems Integration ein System aus verschiedenen Key Performance Indikatoren entworfen und in der Organisation implementiert, das genau das leistet und die Kriterien für CMMI Level 3 erfüllt.
Nowadays, the most employed devices for recoding videos or capturing images are undoubtedly the smartphones. Our work investigates the application of source camera identification on mobile phones. We present a dataset entirely collected by mobile phones. The dataset contains both still images and videos collected by 67 different smartphones. Part of the images consists in photos of uniform backgrounds, especially collected for the computation of the RSPN. Identifying the source camera given a video is particularly challenging due to the strong video compression. The experiments reported in this paper, show the large variation in performance when testing an highly accurate technique on still images and videos.
With the many achievements of Machine Learning in the past years, it is likely that the sub-area of Deep Learning will continue to deliver major technological breakthroughs [1]. In order to achieve best results, it is important to know the various different Deep Learning frameworks and their respective properties. This paper provides a comparative overview of some of the most popular frameworks. First, the comparison methods and criteria are introduced and described with a focus on computer vision applications: Features and Uses are examined by evaluating papers and articles, Adoption and Popularity is determined by analyzing a data science study. Then, the frameworks TensorFlow, Keras, PyTorch and Caffe are compared based on the previously described criteria to highlight properties and differences. Advantages and disadvantages are compared, enabling researchers and developers to choose a framework according to their specific needs.
In this paper we investigate the use of deep neural networks for 3D object detection in uncommon, unstructured environments such as in an open-pit mine. While neural nets are frequently used for object detection in regular autonomous driving applications, more unusual driving scenarios aside street traffic pose additional challenges. For one, the collection of appropriate data sets to train the networks is an issue. For another, testing the performance of trained networks often requires tailored integration with the particular domain as well. While there exist different solutions for these problems in regular autonomous driving, there are only very few approaches that work for special domains just as well. We address both the challenges above in this work. First, we discuss two possible ways of acquiring data for training and evaluation. That is, we evaluate a semi-automated annotation of recorded LIDAR data and we examine synthetic data generation. Using these datasets we train and test different deep neural network for the task of object detection. Second, we propose a possible integration of a ROS2 detector module for an autonomous driving platform. Finally, we present the performance of three state-of-the-art deep neural networks in the domain of 3D object detection on a synthetic dataset and a smaller one containing a characteristic object from an open-pit mine.
In this paper we report on CO2 Meter, a do-it-yourself carbon dioxide measuring device for the classroom. Part of the current measures for dealing with the SARS-CoV-2 pandemic is proper ventilation in indoor settings. This is especially important in schools with students coming back to the classroom even with high incidents rates. Static ventilation patterns do not consider the individual situation for a particular class. Influencing factors like the type of activity, the physical structure or the room occupancy are not incorporated. Also, existing devices are rather expensive and often provide only limited information and only locally without any networking. This leaves the potential of analysing the situation across different settings untapped. Carbon dioxide level can be used as an indicator of air quality, in general, and of aerosol load in particular. Since, according to the latest findings, SARS-CoV-2 can be transmitted primarily in the form of aerosols, carbon dioxide may be used as a proxy for the risk of a virus infection. Hence, schools could improve the indoor air quality and potentially reduce the infection risk if they actually had measuring devices available in the classroom. Our device supports schools in ventilation and it allows for collecting data over the Internet to enable a detailed data analysis and model generation. First deployments in schools at different levels were received very positively. A pilot installation with a larger data collection and analysis is underway.
Daten und Informationen sind die wichtigsten Ressourcen vieler Unternehmen und müssen daher entsprechend geschützt werden. Getrieben durch die erhöhte Vernetzung von Informationstechnologie, die höhere Offenheit infolge datengetriebener Dienstleistungen und eine starke Zunahme an Datenquellen, rücken die Gefahren von Informationsdiebstahl, -manipulation und -verlust in den Fokus von produzierenden Unternehmen. Auf dem Weg zum lern- und wandlungsfähigen Unternehmen kann dies zu einem großen Hindernis werden, da einerseits zu hohe Sicherheitsanforderungen neue Entwicklungen beschränken, andererseits wegen des Mangels an ausreichenden Informationssicherheitskonzepten Unternehmen weniger Innovationen wagen. Deshalb bedarf es individuell angepasster Konzepte für die Bereiche IT-Security, IT-Safety und Datenschutz für vernetzte Produkte, Produktion und Arbeitsplätze. Bei der Entwicklung und Durchsetzung dieser Konzepte steht der Faktor Mensch im Zentrum aller Überlegungen.
In diesem Kapitel wird dargestellt, wie der Faktor Mensch bei der Erstellung von Informationssicherheitskonzepten in verschiedenen Phasen zu beachten ist. Beginnend mit der Integration von Informationssystemen und damit verbundenen Sicherheitsmaßnahmen, über die Administration, bis hin zur Anwendung durch den Endnutzer, werden Methoden beschrieben, die den Menschen, verbunden mit seinem Mehrwert wie auch den Risiken, einschließen. Dabei werden sowohl Grundlagen aufgezeigt als auch Konzepte vorgestellt, mit denen Entscheider in der Unternehmens-IT Leitlinien für die Informationssicherheit festlegen können.
Cybersecurity of Industrial Control Systems (ICS) is an important issue, as ICS incidents may have a direct impact on safety of people or the environment. At the same time the awareness and knowledge about cybersecurity, particularly in the context of ICS, is alarmingly low. Industrial honeypots offer a cheap and easy to implement way to raise cybersecurity awareness and to educate ICS staff about typical attack patterns. When integrated in a productive network, industrial honeypots may not only reveal attackers early but may also distract them from the actual important systems of the network. Implementing multiple honeypots as a honeynet, the systems can be used to emulate or simulate a whole Industrial Control System. This paper describes a network of honeypots emulating HTTP, SNMP, S7communication and the Modbus protocol using Conpot, IMUNES and SNAP7. The nodes mimic SIMATIC S7 programmable logic controllers (PLCs) which are widely used across the globe. The deployed honeypots' features will be compared with the features of real SIMATIC S7 PLCs. Furthermore, the honeynet has been made publicly available for ten days and occurring cyberattacks have been analyzed