The 100 most recently published documents
The present bachelor theses discusses the creation process of a framework for the sys-tematic analysis of twitter posts regarding their sentiment. The result is an application, which links and uses the covered theoretical approaches for text classification.
The task of object detection in the automotive sector can be performed by evaluating various
sensor data. The evaluation of LiDAR data for the detection of objects is a special challenge for
which systems with neural networks can be used. These neural networks are trained by means of a
data set. If you want to use the net with your own recordings or another data set, it is important
to know how well these systems work in combination with data from another sensor. This allows
the results to be estimated in advance and compared with the results of previous experiments.
In this work the sensor dependence of a LiDAR based object recognition with neural networks
will be analysed. The detector used in this work is PointRCNN [1], which was designed for the
KITTI dataset [2]. To check the sensor dependency, the ’AEV Autonomous Driving Dataset’
(A2D2) dataset [3] was selected as a further dataset. After an introduction to PointRCNN and its
functionality, the data of both datasets are analysed. Then the data of the second dataset will be
ported into the format of the KITTI dataset so that they can be used with PointRCNN. Through
experiments with varying combinations of training and validation data it shall be investigated to
what extent trained models can be transferred to other sensor data or datasets. Therefore, it shall
be investigated how strong the dependence of the detector (PointRCNN) on the used sensors is.
The results show that PointRCNN can be evaluated with a different dataset than the training
dataset while still being able to detect objects. The point density of the datasets plays a decisive
role for the quality of the detection. Therefore it can be said that PointRCNN has a sensor
dependency that varies with the nature of the point cloud and its density.
Keywords: LiDAR data, 3D object recognition, laser scanner, sensor dependency, PointRCNN,
PointNet++, PointNet, KITTI Dataset, AEV Autonomous Driving Dataset, A2D2 Dataset
Learning the German language is one of the most critical challenges for refugee children in Germany. It is a prerequisite to allow communication and integration into the educational system. To solve the underlying problem, we conceptualized a set of principles for the design of language learning systems to support collaboration between teachers and refugee children, using a Design Science Research approach. The proposed design principles offer functional and non-functional requirements of systems, including the integration of open educational resources, different media types to develop visual and audio narratives that can be linked to the cultural and social background. This study also illustrates the use of the proposed design principles by providing a working prototype of a learning system. In this, refugee children can learn the language collaboratively and with freely accessible learning resources. Furthermore, we discuss the proposed design principles with various socio-technical aspects of the well-being determinants to promote a positive system design for different cultural and generational settings. Overall, despite some limitations, the implemented design principles can optimize the potential of open educational resources for the research context and derive further recommendations for further research.
Das Ziel der vorliegenden Bachelorarbeit ist die Konzeption eines neuen Ansatzes − die Positive Co-Creation −, der die Elemente des Positive Computing in die Co-Creation integriert. Dafür wurden in einer Literaturanalyse die bestehenden Schwachstellen der Co-Creation herausgearbeitet, um anschließend die Vorteile des Positive Computing aufzuzeigen. Nach der Entwicklung eines spezifischen Modells der Positive Co-Creation, inklusive der verwendeten Methoden und deren Auswirkungen auf die Wohlbefindensfaktoren, wurde das Modell anhand von Experteninterviews evaluiert und verbessert. Das Ergebnis dieser Arbeit ist ein theoretisches Modell der Positive Co-Creation, welches den Prozess vollständig abbildet und einen Ansatzpunkt für eine praktische Umsetzung bildet. Dieser Ansatz ist gut geeignet, um bestehende Co-Creation-Prozesse anhand von Technologien um die Aspekte des Wohlbefindens zu erweitern.
The highly successful lecture series on the topic of measurement and sensor technologies as part of the IEEE Workshop at the University of Applied Sciences Ruhr West (HRW) is being continued in collaboration with the University of Siegen, the TU Chemnitz and the ITMO National Research University of Information Technologies, Mechanics and Optics in St. Petersburg. This time the event is featuring an even more international orientation by linking it with the Russian SENSORICA. The topics cover industrial and medical measurement technology as well as sensor technology in vehicles. Our event offers a platform for knowledge transfer between industry and public and commercial research institutions in the area of measurement technology.
This Abstract Book offers the opportunity of contacting speakers even after the event.
In addition we are very pleased to have selected contributions published in a special edition of the journal „tm Technisches Messen“ (De Gruyter Oldenbourg Verlag) again this year.
The detection of soil erosion processes in dams, hydraulic heave failure or corrosion processes of reinforcing steel in concrete are a small selection of measuring applications in civil engineering where the impedance analysis can be used to determine the measurand. Those measuring applications are having high requirements for the measuring hardware. For example a common interface for fast data exchange, high resolution, independent functionality and easy customizability to suit the measuring application. For that reason, a well-known application for steel-mill process monitoring can be used as a development platform. This hardware platform is based on a vector network analyzer and is meeting the requirements mainly. However, a couple of modifications has to be made, like replacing the ADC for a higher sample rate, Ethernet for easy and fast data exchange and the microcontroller for more calculation power.
Process Monitoring in Steel-Mills using Impedance Analysis: VNA Improvement for Data Acquisition
(2017)
The process automation extends over every manufacturing step of a product in the steel-mill to increase the quality, quantity and energy efficiency. The product dimensions are an important part of the quality control; these must maintain the specified tolerances. Additional to the cross-sectional-area, the measured data contains much more information about the manufacturing process, e.g. eccentricity, condition of the rolls and defects of the rod. For analyzing the measured data and to gather more information about the manufacturing process it is necessary to increase the speed of the data acquisition by performing some modifications of the VNA, e.g. faster analog to digital converter and microcontroller, improved firmware and optimized values of the passive electrical components for faster time constants and transient responses.
Rolling mills are continually improved and opti-mized by implementing innovative technology to decrease costs and scrap. Despite of the progressive automation and experience, some important process parameters can still not be determined with sufficient accuracy. As part of the research project PIREF, the velocity of the hot rolled rod shall be measured by using im-pedance analysis to estimate the volumetric flow rate of the mate-rial. For a high accuracy measurement of the impedance, a pow-erful VNA is used. To minimize errors in the measurement, caused by e.g. temperature drift, a correction of the measurement fre-quency is needed. This must be achieved without recalibration of the VNA to avoid faulty behavior of the online control. To solve this problem, an approach based on a polynomial regression is presented in this work.
Quality and dimensional accuracy of hot rolled steel rods depend on several process parameters. In fact many of these crucial parameters are not be sufficiently determined yet. By improving automation and process control costs and scrap of production can be decreased. As part of the research project PIREF, one of these parameters – the roll gap – is under investigation beside other topics. Before starting rolling, the roll gap is typically set to a fixed value according to the planed dimensions of the product, but the forces during the rolling of the rod cause an enlargement of the roll gap. In which way the rolls change their position and form shall be examined in our research project. Therefore a first experimental setup has been built up to determine the change in position of the rolls under applied force. This is realized by a pot core coil as sensor using impedance analysis. The first results are presented in this work as a proof-of-principle.
Process diagnosis is an important method for improving product quality in rolling mills. In addition, the measurement of process variables such as roll gap, cross-sectional area, velocity, and volume flow of the material during production enables the implementation of model-based control concepts to improve product quality. The non-contact speed measurement of hot wire and bar is still a big challenge due to the rough environmental conditions and is solved mainly with optical measuring methods in production. The alternative measurement principle with eddy current sensors presented in this paper enables velocity measurement at locations in a rolling mill where optical measurement methods are not suitable.
In the field of producing hot-rolled steel bars and wires, hot rolling mills are incomplete or barely equipped with measuring technology for recording relevant process parameters. Therefore, there is a big potential to increase product quality and to decrease costs and scrap by improving process control establishing new sensor systems. One of these crucial parameters is the roll gap,which is investigated as part of the research project PIREF. In this paper an experimental setup for examining the roll gap during a rolling process is presented and based on these results different sensor arrangements are discussed.
Velocity Approximation of Hot Steel Rods Using Frequency Spectroscopy of the Cross-Section Area
(2019)
In this work, an approach for velocity approximation of hot steel rods based on frequency spectroscopy is presented. For this purpose, a sensor already implemented in a rolling mill for measuring the cross-sectional area of the rolling stock is used to obtain information about the velocity of the hot rods. Moreover, the effect of forward slip is briefly discussed.
The development of innovative measuring technology for process optimization in hot rolling mills becomes more and more relevant because of increasing demands on product quality. Measurement technology for high-resolution non-contact cross-sectional area measurement has shown that the variation in cross-sectional area contains information about the rolling process. This information can be used for the development of new measurement devices and analytical methods for process optimization. The harsh environmental conditions and strict safety regulations result in great effort when implementing a new sensor prototype in hot rolling mills. For this reason, this work presents a mechatronic test stand that can simulate the cross-sectional area variation under laboratory conditions realistically.
Researchers have previously utilized the advantages of a design driven by well-being and intergenerational collaboration (IGC) for successful innovation. Unfortunately, scant information exists regarding barrier dimensions and correlated design solutions in the information systems (IS) domain, which can serve as a starting point for a design oriented toward well-being in an IGC system. Therefore, in this study, we applied the positive computing approach to guide our analysis in a systematic literature review and developed a framework oriented toward well-being for a system with a multi-generational team. Our study contributes to the IS community by providing five dimensions of barriers to IGC and the corresponding well-being determinants for positive system design. In addition, we propose further research directions to close the research gap based on the review outcomes.
Globalization and information technology enable people to join the movement of global citizenship and work without borders. However, different type of barriers existed that could affect collaboration in today’s work environment, in which different generations are involved. Although researchers have identified several technical barriers to intergenerational collaboration (iGOAL), the influence of cultural diversity on iGOAL has rarely been studied. Therefore, using a quantitative study approach, this paper investigates the impact of differences in cultural background on perceived technical and operational barriers to iGOAL. Our study reveals six barriers to IGC that are perceived differently by culturally diverse people (CDP) and non-CDP. Furthermore, CDP can foster IGC because CDP consider the barriers to be of less of a reason to avoid working with different generations than do non-CDP.
Enabling decentral collaborative innovation processes -a web based real time collaboration platform
(2018)
The main goal of this paper is to define a collaborative innovation process as well as a supporting tool. It is motivated through the increasing competition on global markets and the resultant propagation of decentralized projects with a high demand of innovative collaboration in global contexts. It bases on a project accomplished by the author group. A detailed literature review and the action design research methodology of the project led to an enhanced process model for decentral collaborative innovation processes and a basic realization of a browser based real time tool to enable these processes.The initial evaluation in a practical distributed setting has shown that the created tool is a useful way to support collaborative innovation processes.
Vor dem Hintergrund globaler Vernetzung, Digitalisierung und Big Data geht das Excel-Zeitalter seinem Ende entgegen. Neue Technologien versprechen eine höhere Effizienz und Effektivität des Controllings. Die Veränderungen umfassen dabei alle Aufgabenbereiche des Controllings – allerdings mit unterschiedlicher Geschwindigkeit und differenzierten Lösungsansätzen. Der vorliegende Beitrag untersucht drei Technologien, die im Controlling derzeit besonders stark diskutiert werden: Robotic Process Automation, Predictive Analytics und Künstliche Intelligenz. Im Rahmen eines Forschungsseminars an der Hochschule Ruhr West in Mülheim a.d.R. wurden dazu 57 Führungskräfte aus Controlling und Rechnungswesen befragt. Über 75% der befragten Unternehmen halten die Digitalisierung für wichtig, mehr als die Hälfte (52%) beabsichtigen, eine oder mehrere der drei untersuchten Technologien in den nächsten drei Jahren neu einzusetzen. In den Aufgabenbereichen ist bislang vor allem das Reporting digital unterstützt, in der Strategischen Planung und im Risikomanagement ist dagegen die Digitalisierung gering. Wesentliche Hindernisse sehen die Befragten im fehlenden Know-How und mangelnder Veränderungsbereitschaft.
Open Educational Resources (OER) intend to support access to education for everyone. However, this potential is not fully exploited due to various barriers in the production, distribution and the use of OER. In this paper, we present requirements and recommendations for systems for global OER authoring. These requirements as well as the system itself aim at helping creators of OER to overcome typical obstacles such as lack of technical skills, different types of devices and systems as well as the cultural differences in cross-border-collaboration. The system can be used collaboratively to create OER and supports multi-languages for localization. Our paper contributes to facilitate global, collaborative e-Learning and design of authoring platforms by identifying key requirements for OER authoring in a global context.
DamokleS 4.0
(2019)
Dieser interne Bericht beschreibt die Zielsetzung, Durchführung und Auswertung des Projektes Damokles 4.0. Das Projekt zielt darauf ab, neue, digitale Technologien in die Schwerindustrie einzuführen um Produktionsprozesse zu modernisieren. Unter Einsatz neuer Technologien, insbesondere mobiler Geräte, soll ein cyberphyiskalisches System (CPS) eine kontextbasierte und künstlich intelligente Unterstützung der Mitarbeiter in den Bereichen der Schwerindustrie ermöglichen. Hierzu werden typische Anwendungsfälle und die damit verbundenen Szenarien zur Unterstützung der Mitarbeiter auf Basis von neuen, flexiblen, adaptiven und mobilen Technologien, wie Augmented Reality und künstlicher Intelligenz, modelliert. Um den Prototypen einer AR-Anwendung und einer kamerabasierte Personenverfolgung zu entwickeln, hat die Hochschule Ruhr West im kleinen Technikum am Campus Bottrop eine entsprechende industrielle Umgebung simuliert. Die Projektergebnisse zeigen die Anwendbarkeit der vorgeschlagenen Softwareansätze und die Ergebnisse einer Untersuchung der psychologischen Einflüsse auf die Mitarbeiter.
Digital transformation is a process of digitizing the working and living environment in which people are at the center of digitization. In this paper, we present a personas-based guideline for system developers on how the humanization of digital transformation integrates into the design process. The proposed guideline uses the positive personas from the beginning as a basis for the transformation of the working environment into the digital form. We used the literature research as a preliminary study for the process of wellbeing-driven digital transformation design, consisting of questions for structuring the required information in the positive personas as well as a potential method that could be integrated into the wellbeing-based design process.
Why Should the Q-method Be Integrated Into the Design Science Research? A Systematic Mapping Study
(2019)
The Q-method has been utilized over time in various areas, including information systems. In this study, we used a systematic mapping to illustrate how the Q-method was applied within Information Systems (IS) community and proposing towards integration of Q-method into the Design Sciences Research (DSR) process as a tool for future research DSR-based IS studies. In this mapping study, we collected peer-reviewed journals from Basket-of-Eight journals and the digital library of the Association for Information Systems (AIS). Then we grouped the publications according to the process of DSR, and different variables for preparing Q-method from IS publications. We found that the potential of the Q-methodology can be used to support each main research stage of DSR processes and can serve as the useful tool to evaluate a system in the IS topic of system analysis and design
The virtual classroom continues to grow, but it is becoming more and more the norm, and it is fundamentally different from the vocational students at the Indonesian university. With the promised benefits of the virtual classroom, many challenges and difficulties come in the implementation. Although there are already successful design principles for virtual classrooms that support organizations in overcoming the challenges, the approach to implementing the design principles of virtual classroom at the vocational higher education in Indonesia is still lacking. In this study, we aim to answer the research gap and used the design sciences research by interviewing the lecturers to design the solutions. The proposed design approaches were implemented in a course and evaluated with students from two different groups. Overall, the evaluation of the proposed approaches shows1 significant results as an indicator of the benefits of the implementation of a virtual classroom for vocational students in Indonesia.
In diesem Artikel wird ein System vorgestellt, welches eine videobasierte Hinderniserkennung zur automatisierten Bildanalyse von Straßenverkehrsszenen durchführt. Eine Unterteilung der Hinderniserkennung in Objektdetektion, Objektverfolgung und Objektklassifikation lässt eine Extraktion und eine Attributierung von Verkehrsteilnehmern zu. Eine Szeneninterpretation ist ableitbar.
Positive Computing umfasst Design, Realisierung und Bewertung von Anwendungssystemen und deren Einflüsse mit dem Ziel, Lebensqualität und Wohlbefinden von Menschen zu verbessern und sie bei der Entfaltung ihrer Potenziale zu unterstützen. Das Institut Positive Computing (IPCo) an der Hochschule Ruhr West soll dieses neue Paradigma in einem inter- und transdisziplinären Ansatz erschließen, untersuchen und umsetzen. Das Paradigma ist anwendbar auf nahezu alle Bereiche des privaten und beruflichen Lebens. Die Forschung des IPCo fokussiert zunächst jedoch auf die positive Nutzung von Informations- und Kommunikationstechnologien (IKT) für generationenübergreifende Herausforderungen. Hierzu sollen technologische Lösungen unter kontinuierlicher Einbeziehung menschlicher Bedürfnisse und sozialer Fragestellungen erarbeitet
werden.
The adoption of Open Educational Resources (OER) can support collaboration and knowledge sharing. One of the main areas of the usage OER is the internationalization, i.e., the use in a global context. However, the globally distributed co-creation of digital materials is still low. Therefore, we identify essential barriers, in particular for co-authoring of OER in global environments. We use a design science research method to introduce a barrier framework for co-authoring OER in global settings and propose a wellbeing-based system design constructed from the barrier framework for OER co-authoring tool. We describe how positive computing concepts can be used to overcome barriers, emphasizing design that promotes the author's sense of competence, relatedness, and autonomy.
Technologie die beflügelt
(2016)
Das CameraFramework wurde entwickelt, um mittels Socket-Kommunikation [1] als Middleware zwischen verschiedenen Kamerainstanzen mit eigenen Kameratreibern und Clienten zu fungieren. Über diesen Kommunikationsweg ist es möglich Clienten nicht nur lokal, sondern auch über das Netzwerk mit Kameradaten zu versorgen. Um neue Kameras mit dem Framework nutzen zu können, muss die Implementierung gewissen Regeln folgen, was durch ein vorgegebenes Basis-Interface (abstrakte Basis-Klasse in C++ [2]) fast vollständig sichergestellt ist. Neue Kameras werden zur Laufzeit über dynamische Bibliotheken geladen. Parameter für Kameras sind über ein XML-File [3] einzustellen. Funktionen zur Übergabe von neuen Kameradaten sind implementiert und müssen durch den Entwickler der einzelnen Kamerainterfaces aufgerufen werden.
Die Zuordnung von Kameradaten zum passenden Nutzer übernimmt das Framework. Jeder Clienterhält seinen eigenen konfigurierbaren Ringbuffer [4] um unabhängig von anderen Nutzern und Kameras zu sein. Die Aufgaben des Frameworks sind auf verschiedene Module, wie in Abbildung 1 dargestellt, aufgeteilt.
Autonomous driving is one of the future visions in which many vehicle manufacturers are working with high pressure.
Nowadays, it is already supported partially by high-class vehicles. A completely autonomous journey is indeed the goal, but in cars for
the public road traffic still not available. Automatic lane keeping assistants, speed regulators as well as shield and obstacle detections
are parts or precursors on the way to completely autonomous driving.
The American vehicle manufacturer Tesla is not only known for its electric drive, but also for the fact that high-pressure work is carried out on the autonomous drive. Tesla is thus the only vehicle manufacturer to use its users as so-called beta testers for its assistance systems. The progress and the function of the currently available Model S in the field of assistance systems and autonomic driving is documented and described in this paper. It is shown how good or bad the test vehicle manages scenarios in normal road traffic situations
with the assistance systems, e.g. lane keeping assistant, speed control, lane change and distance assistant, and which scenarios can
not be managed by the vehicle itself.
Systeme zur automatisierten Bildanalyse sind vielfältig einsetzbar und gewinnen aufgrund technologischer Weiterentwicklungen und gesellschaftlicher Akzeptanz zunehmend an Bedeutung. Schwerpunkt im Bereich der "Technischen Bildverarbeitung dynamischer Szenen" ist die Entwicklung von Methoden, die bei der Interpretation von Bildern aus verschiedenen Sensordaten Verwendung finden. Dies sind neben den herkömmlichen Kamerabildern im wesentlichen Röntgen- und Radarbilder. Unter geeigneter Berücksichtigung der durch die jeweiligen Anwendungen vorgegebenen Randbedingungen werden daraus entsprechende Verfahren abgeleitet. Derzeitige Projekte beschäftigen sich mit der Analyse von Straßenverkehrsszenen, der Detektion von Sprengstoffzündern bei der Durchleuchtung von Fluggepäck, sowie mit der Bestimmung von Art und Ausdehnung von Ölverschmutzungen bei der Meeresüberwachung.
We are “not” too (young/old) to collaborate: Prominent Key Barriers to Intergenerational Innovation
(2019)
In this study, we analyzed the barriers to technology-supported intergenerational innovation to understand better how young and old can collaborate towards global innovations. Researchers in different disciplines have already identified various barriers to intergenerational collaboration. However, barriers are changing depending on the context of collaboration, and difficulties still exist to support intergenerational innovation in global settings. Therefore, we investigated the barriers that emerge when people work with someone decades older or younger. The results of our study have shown what barriers are influenced by age, what barriers exist only for senior and younger adults. The study theoretically contributes to deepening the Information Systems (IS) community's understanding of the barriers to intergenerational innovation that need to be considered when developing systems for global innovation
Systems for automated image analysis are useful for a variety of tasks and their importance is still increasing due to technological advances and an increase of social acceptance. The main focus of "Technical Image Processing of Dynamic Scenes" lies
with the development of methods for the interpretation of images derived from various sensors. Apart from conventional visual images, this involves mainly X-ray and radar images. Taking into account the requirements of the various applications, suitable methods are derived. Current projects are dealing with the analysis of traffic scenes, detection of detonators when X-raying luggage and determination of type and expansion of oil pollution in maritime surveillance.
Aktiv im Alter
(2016)
Die Prognosen für den demografischen Wandel sind eindeutig: In den kommenden Jahren wird es immer mehr Menschen über 65 Jahre geben. Damit verbunden sind große Herausforderungen für die Gesellschaft und ihre Sozialsysteme, aber auch für viele Angehörige, die ihre Verwandten im Alter pflegen. Doch nicht alle älteren Menschen leben im Kreise ihrer Familie oder können sich Fremdbetreuung durch Pflegedienste leisten. Häufig übernehmen Nachbarn oder Freunde aus der Umgebung diese Aufgabe. Für diese Menschen wird das Wohnquartier zum zentralen Gesundheitsstandort.
Im besten Fall können sie dort ihren Alltag noch lange selbstständig bewältigen und ihre sozialen Kontakte aufrechterhalten. Das soll bald eine App unterstützen. Sie ist Teil eines Trainingsprogramms, das die Hochschule für Gesundheit (hsg) im Verbund mit der Hochschule Ruhr West erarbeitet. Der Name des Projekts ist Programm: „Quartier agil – Aktiv vor Ort“. Mit Übungen zum kognitiven und körperlichen Training, Angeboten für Gruppenaktivitäten, Kommunikationsforen und Funktionen zur Selbstkontrolle wollen die Forscherinnen und Forscher
ältere Menschen fit halten.
In der vorliegenden Arbeit wird ein Verfahren vorgestellt, welches textur- und konturbasierte Verfahren zur Segmentierung fusioniert. Als Kopplungsmatrix wird eine selbstorganisierte Karte nach Kohonen verwendet. Eine verbesserte Objekt- zu Hintergrundtrennung im Vergleich zu Einzelalgorithmen wird demonstriert.
Für das sichere Führen von Fahrzeugen im Straßenverkehr ist ein hohes Maß an Informationsverarbeitung notwendig, um aus den zur Verfügung stehenden Informationen, Handlungen für die Fahrzeugsteuerung abzuleiten. Der Mensch löst diese Aufgabe hauptsächlich auf der Basis visueller Informationen. Durch die Arbeitsweise des menschlichen Gehirns motiviert, wird am Institut für Neuroinformatik der Ruhr-Universität Bochum an einer Fahrzeugführung mittels Computer Vision gearbeitet. Fortlaufend oder zumindest in kurzen Abständen müssen hierbei Verkehrsteilnehmer aus den visuellen Informationen extrahiert und danach weiter attributiert werden. Wichtige Eigenschaften sind hierbei: Objektklasse (PKW, LKW, Fußgänger etc.), Abstand, Geschwindigkeit, Bewegungsrichtung und das Gefahrenpotential bezüglich der eigenen Ortsveränderung. Die Vielzahl der durch die Umwelt aufgestellten Randbedingungen und das aus der Aufgabenstellung implizierte hohe Maß an Sicherheit bedingen ein robustes und flexibles Gesamtsystem. Dieses Gesamtsystem besteht zum einen aus Basis-Algorithmen zur Vorverarbeitung der Eingabedaten und Extraktion von Bildmerkmalen und zum anderen aus darauf aufbauenden Verfahren zur Segmentierung, Klassifizierung und Verfolgung von Fahrzeugen.
Technical Report
(2016)
This internal report discusses the theoretical and practical aspects of the cluster management framework SimpleHydra, which was developed in order to allow researchers the quick setup of classical small to mid-scale computation clusters while being as lightweight and platform independent as possible. We motivate crucial design choices with a theoretical analysis in the aspect of time and space complexity, furthermore we give a comprehensive introduction regarding the frameworks usage (which includes examples and detailed description of fundamental concepts as well as data structures). In addition to that we illustrate application scenarios with complete source code examples. Furthermore we hope that this document proves valuable not only as a development report but also as a practical manual for SimpleHydra.
We present a study on 3D based hand pose recognition using a new generation of low-cost time-of-flight(ToF) sensors intended for outdoor use in automotive human-machine interaction. As signal quality is impaired compared to Kinect-type sensors, we study several ways to improve performance when a large number of gesture classes is involved. We investigate the performance of different 3D descriptors, as well as the fusion of two ToF sensor streams. By basing a data fusion strategy on the fact that multilayer perceptrons can produce normalized confidences individually for each class, and similarly by designing information-theoretic online measures for assessing confidences of decisions, we show that appropriately chosen fusion strategies can improve overall performance to a very satisfactory level. Real-time capability is retained as the used 3D descriptors, the fusion strategy as well as the online confidence measures are computationally efficient.
Multimodaler Sensor zur Fahrzeugführung: Teilprojekt: Architektur, Rundumsicht und Objekterkennung
(1997)
Analyse dynamischer Szenen
(1999)
In diesem Artikel wird die Analyse dynamischer Szenen im Rahmen einer flexiblen Architektur zur Lösung von Fahrerassistenzaufgaben in Kraftfahrzeugen vorgestellt. Die Lösung unterschiedlicher Aufgaben mit verwandten Ansätzen bedingt einen hohen Grad an Modularität und Flexibilität. Nur so können die gestellten Aufgaben mit den vorhandenen Algorithmen optimal gelöst werden. In der vorgestellten Architektur wird eine objektbezogene Analyse von Sensordaten, eine verhaltensbasierte Szeneninterpretation und eine Verhaltensplanung durchgeführt. Eine globale Wissensbasis, auf der jedes einzelne Modul arbeitet, beinhaltet die Beschreibung physikalischer Zusammenhänge, Verhaltensregeln für den Straßenverkehr, sowie Objekt- und Szenenwissen.
Externes Wissen (z.B. GPS – Global Positioning System) kann ebenfalls in die Wissensbasis eingebunden werden. Als Anwendungsbeispiel der Verhaltensplanung ist ein intelligenter Tempomat realisiert.
For face recognition from video streams speed and accuracy are vital aspects. The first decision whether a preprocessed image region represents a human face or not is often made by a feed-forward neural network (NN), e.g. in the Viisage-FaceFINDER® video surveillance system. We describe the optimisation of such a NN by a hybrid algorithm combining evolutionary multi-objective optimisation (EMO) and gradient-based learning. The evolved solutions perform considerably faster than an expert-designed architecture without loss of accuracy. We compare an EMO and a single objective approach, both with online search strategy adaptation. It turns out that EMO is preferable to the single objective approach in several respects.
Das kEFIR‐Projekt untersucht die praktische Anwendung von thermographischen Verfahren zur Analyse der strukturellen Integrität von Windkraftrotorblättern. Das Projekt entstand in Zusammenarbeit der Hochschule Ruhr West (HRW) mit der IQbis Consulting GmbH im Rahmen eines ZIM‐Förderprojekts des Bundesministeriums für Wirtschaft und Energie (BMWi). Hintergrund ist die zunehmende Anzahl von Windkraftanlagen (WKA) und der somit steigende Wartungsaufwand. Um einen reibungslosen Betrieb dieser Anlagen zu gewährleisten und damit den besonderen Anforderungen an die Verfügbarkeit energieerzeugender Anlagen sicherzustellen, ist ein Bedarf an qualitativ hochwertigen Fehleranalysesystemen für im Betrieb befindlicher WKA von besonderer Bedeutung. Erfahrungsgemäß ist der Zeitaufwand für diese Inspektionen mit aktuellen Mitteln sehr groß und wird üblicherweise mit mehreren Arbeitstagen kalkuliert. Die Reproduzierbarkeit der gewonnenen Daten ist bei den derzeitigen Methoden meist nicht gewährleistet. Um frühzeitig auf Instabilitäten oder Schäden in den Rotorblättern einer WKA aufmerksam zu werden, ist die Entwicklung eines schnellen und qualitativ hoch wertigen Fehleranalysesystems von zentraler Bedeutung. Ein Forschungsschwerpunkt in diesem Zusammenhang ist die Entwicklung von geeigneten bildgebenden und berührungslosen Verfahren, welche bei den Inspektionen eingesetzt werden können. Beispielsweise erlaubt der Einsatz thermographischer Sensoren eine Analyse nicht nur der Rotorblattoberfläche, sondern auch ihrer inneren Struktur. Weiterhin ist aufgrund des schnell wachsenden Marktes bei unbemannten Luftfahrzeugen, wie beispielsweise positionsstabiler Quatrocoptersysteme, eine zusätzliche Möglichkeit gegeben, die Inspektion von Windenergieanlagen mit Hilfe mobiler, kompakter und fliegender Analysesysteme zu unterstützen.
Das kEFIR‐Projekt untersucht die praktische Anwendung von thermographischen Verfahren zur Analyse der strukturellen Integrität von Windkraftrotorblättern. Das Projekt entstand in Zusammenarbeit der Hochschule Ruhr West (HRW) mit der IQbis Consulting GmbH im Rahmen eines ZIM‐Förderprojekts des Bundesministeriums für Wirtschaft und Energie (BMWi). Hintergrund ist die zunehmende Anzahl von Windkraftanlagen (WKA) und der somit steigende Wartungsaufwand. Um einen reibungslosen Betrieb dieser Anlagen zu gewährleisten, und damit den besonderen Anforderungen an die Verfügbarkeit energieerzeugender Anlagen sicherzustellen, ist ein Bedarf an qualitativ hochwertigen Fehleranalysesystemen für im Betrieb befindlicher WKA von besonderer Bedeutung. Erfahrungsgemäß ist der Zeitaufwand für diese Inspektionen mit aktuellen Mitteln sehr groß und wird üblicherweise mit mehreren Arbeitstagen kalkuliert. Die Reproduzierbarkeit der gewonnenen Daten ist bei den derzeitigen Methoden meist nicht gewährleistet. Um frühzeitig auf Instabilitäten oder Schäden in den Rotorblättern einer WKA aufmerksam zu werden, ist die Entwicklung eines schnellen und qualitativ hochwertigen Fehleranalysesystems von zentraler Bedeutung. Ein Forschungsschwerpunkt in diesem Zusammenhang ist die Entwicklung von geeigneten bildgebenden und berührungslosen Verfahren, welche bei den Inspektionen eingesetzt werden können. Beispielsweise erlaubt der Einsatz thermographischer Sensoren eine Analyse nicht nur der Rotorblattoberfläche, sondern auch ihrer inneren Struktur. Weiterhin ist aufgrund des schnell wachsenden Marktes bei unbemannten Luftfahrzeugen, wie beispielsweise positionsstabiler Quatrocoptersysteme, eine zusätzliche Möglichkeit gegeben, die Inspektion von Windenergieanlagen mit Hilfe mobiler, kompakter und fliegender Analysesysteme zu unterstützen.
Es ist eine alltägliche Erfahrung, daß wir Urteile über gut oder schlecht, bzw. qualitativ hochwertig oder minderwertig eines Gegenstandes mit der Wahrnehmung des emittierten Geräuschschalls in Verbindung bringen. Der Geräuschlaut ist deshalb ein wichtiges Entscheidungskriterium bei der Auswahl eines Produktes, welches wahrnehmbaren Schall erzeugt. Die Fragestellung hinsichtlich der Geräuschqualität und des Geräuschdesigns stellt daher hohe Anforderungen an den Akustik-Ingenieur. Zum heutigen Zeitpunkt ist es jedoch nicht möglich, mit einer instrumentellen Meßtechnik Aussagen über die Eignung eines Geräuschschalls für ein Produkt zu machen. Es ist nicht möglich, kognitive Faktoren über eine instrumentelle Meßtechnik zu messen. Es reicht nicht aus, eine Geräuschgüte mit Bewertungsschemata wie dem A-bewerteten Schalldruckpegel
oder Lautheitsmodellen zu definieren. Diese lassen allein keine eindeutigen Aussagen über die Wahrnehmung von Geräuschen zu. Der vorliegende Beitrag ist als Ansatz für das Soundengineering von Fahrzeuginnengeräuschen zu sehen. Es wird anhand von Hörversuchen mit Fahrzeuginnengeräuschen ein objektiver Beschreibungskatalog ermittelt, der eine Aussage über die jeweilige Hörempfindung zuläßt.
We propose a new approach to object detection based on data fusion of texture and edge information. A self organizing Kohonen map is used as the coupling element of the different representations. Therefore, an extension of the proposed architecture incorporating other features, even features not derived from vision modules, is straight forward. It simplifies to a redefinition of the local feature vectors and a retraining of the network structure. The resulting hypotheses of object locations generated by the detection process are finally inspected by a neural network classifier based on co-occurence matrices.
Systems for automated image analysis are useful for a variety of tasks and their importance is still increasing due to technological advances and an increase of social acceptance. Especially in the field of driver assistance systems the progress in science has reached a level of high performance. Fully or partly autonomously guided vehicles, particularly for road-based traffic, pose high demands on the development of reliable algorithms due to the conditions imposed by natural environments. At the Institut fur Neuroinformatik, methods for analyzing driving relevant scenes by computer vision are developed in cooperation with several partners from the automobile industry. We introduce a system which extracts the important information from an image taken by a CCD camera installed at the rear view mirror in a car. The approach consists of a sequential and a parallel sensor and information processing. Three main tasks namely the initial segmentation (object detection), the object tracking and the object classification are realized by integration in the sequential branch and by fusion in the parallel branch. The main gain of this approach is given by the integrative coupling of different algorithms providing partly redundant information.
Im vorliegenden Beitrag wird ein hochsprachenprogrammierbares System zur schritthaltenden Vollbild-Interpretation natürlich beleuchteter Szenenfolgen im Videotakt vorgestellt. Im einzelnen werden folgende Teilmodule und Subsysteme beschrieben: eine hochdynamische, pixellokal autoadaptive CMOS-Kamera mit ca. 120 dB Helligkeitsdynamik (20Bits/Pixel) ein hochsprachenprogrammierbarer Systolic Array Prozessor (für die pixelbezogenen Verarbeitungsmodule) im PCI-Kartenformat, samt optimierendem Compiler, Simulator und Emulator Systemprozeßgerüste unter Linux auf den für die Echtzeit-Anwendungen eingesetzten Hostrechnern (z.B. DEC/Alpha oder Intel/ Pentium)eine prototypische Anwendung zur bildverarbeitungsbasierten Eigenbewegungsbeobachtung (Translationsrichtung, Eotationsraten)eine prototypische, automotive Anwendung zur schritthalt enden Detektion und Kartierung des Straßen- und Spurverlaufs unter partieller monokularer 3D-Rekonstruktion, sowie prototypische Anwendungen zur Klassifikation verkehrsrelevanter Hindernisse (Verkehrsteilnehmer)
Das vorliegende Paper gibt einen Überblick über das Verhalten von modernen, autonom navigierenden Fahrzeugen in Baustellen. Dabei werden besondere Herausforderungen für die autonome Navigation im Baustellenbereich benannt. Außerdem wird ein Überblick über die Sensorausstattung und die Fahrerassistenzsysteme von modernen Fahrzeugen gegeben und es werden Technologien vorgestellt, die für eine Verbesserung der autonomen Navigation durch Baustellen genutzt werden können. Es wird ein Versuch durchgeführt, der aufzeigt, wie zuverlässig moderne Fahrzeuge durch Baustellensituationen navigieren können. Dabei werden Schwachstellen, wie bspw. die mangelnde Verfügbarkeit von Fahrerassistenzsystemen bei niedrigen Geschwindigkeiten, aufgedeckt.
Derzeitige Projekte am Institut für Neuroinformatik in Bochum beschäftigen sich mit der Analyse von Straßenverkehrsszenen mittels Computer Vision [12]. Dies impliziert, wegen der durch die natürliche Umwelt aufgestellten Randbedingungen, hohe Anforderungen an die zu entwickelnden Algorithmen. Im speziellen wird versucht, Verkehrsteilnehmer aus Videobildern zu extrahieren und die so gewonnenen Objekthypothesen weiter zu attributieren (z.B. Objektklasse, Abstand, Geschwindigkeit, Gefahrenpotential hinsichtlich der beabsichtigten Eigentrajektorie etc.), um im Hinblick auf den Einsatz in Assistenzsystemen in Fahrzeugen eine möglichst genaue Beschreibung der Umwelt zu erreichen. Nicht nur die große Vielfalt der unterschiedlichen Umweltszenarien, sondern auch das hohe Maß an Sicherheit, das die gestellte Aufgabe erfordert, bedingen ein breitbandiges und flexibles Gesamtsystem [6]. Ein Lösungsvorschlag wird im folgenden behandelt.
Analysis of dynamic scenes
(2000)
In this paper the proposed architecture for a dynamic scene analysis is illustrated by a driver assistance system. To reduce the number of traffic accidents and to increase the drivers comfort, the thought of designing driver assistance systems rose in the past years. Principal problems are caused by having a moving observer (ego motion) in predominantly natural surroundings. In this paper we present a solution for a flexible architecture for a driver assistance system. The architecture can be subdivided into four different parts: the object-related analysis, the knowledge base, the behavior-based scene interpretation, and the behavior planning unit. The object-related analysis is fed with data by the sensors (vision, radar). The sensor data are preprocessed (flexible sensor fusion) and evaluated (saliency map) searching for object-related information (positions, types of objects, etc.). The knowledge base is represented by static and dynamic knowledge. It consists of a set of rules (traffic rules, physical laws), additional information (GPS, lane-information) and it is implicitly used by algorithms in the system. The scene interpretation combines the information extracted by the
object-related analysis and inspects the information for contradictions. It is strongly connected to the behavior planning using only information needed for the actual task. In the scene interpretation consistent representations (i.e., bird’s eye view) are organized and interpreted as well as a scene analysis is performed. The results of the scene interpretation are used for decision making in behavior planning, which is controlled by the actual task.
Handgesten im Automobil haben das Potenzial einer Kombination von gut sichtbaren Displays nahe der Windschutzscheibe und einer als intuitiv empfundenen Gestensteuerung, wie sie berührungsgesteuert von Smartphones aber auch berührungslos von einigen Fernsehgeräten bekannt ist. Bei entsprechender Positionierung der Sensoren können so die Augen auf der Straße und die Hände am Lenkrad oder zumindest sehr nahe dazu verbleiben. Der hier beschriebene frühe Demonstrator zeigt die Machbarkeit dieser Technologie mit einem neuartigen Erkennungsverfahren.
Ziel des Verbundprojektes APFel (Projektlaufzeit: 01.01.2010 ‐ 31.03.2014)war eine zeitlich vorwärts‐ und rückwärtsgerichtete Lokalisation von Personen innerhalb eines Kameranetzwerkes aus sich nicht überlappenden Kameras in Hyperechtzeit zu ermöglichen. Einsatzbereiche dieses Szenarios sind kritische Infrastrukturen wie Flughäfen und Flugplätze. Zunächst fokussierte das Projekt APFel auf die Lokalisation einer einzelnen Zielperson. Weiterführend wurden die entwickelten Verfahren auf die Analyse von Gruppen erweitert, um Personen als Teil einer Gruppe lokalisieren zu können.
We present a novel approach of distributing matrix multiplications among GPU-equipped nodes in a cluster system. In this context we discuss the induced challenges and possible solutions. Additionally we state an algorithm which outperforms optimized GPU BLAS libraries for small matrices. Furthermore we provide a novel theoretical model for distributing algorithms within homogeneous computation systems with multiple hierarchies. In the context of this model we develop an algorithm which can find the optimal distribution parameters for each involved subalgorithm. We provide a detailed analysis of the algorithms space and time complexities and justify its use with a structured evaluation within a small GPU-equipped Beowulf cluster.
We present a novel method to perform multi-class pattern classification with neural networks and test it on a challenging 3D hand gesture recognition problem. Our method consists of a standard one-against-all (OAA) classification, followed by another network layer classifying the resulting class scores, possibly augmented by the original raw input vector. This allows the network to disambiguate hard-to-separate classes as the distribution of class scores carries considerable information as well, and is in fact often used for assessing the confidence of a decision. We show that by this approach we are able to significantly boost our results, overall as well as for particular difficult cases, on the hard 10-class gesture classification task.
A light-weight real-time ap- plicable hand gesture recognition system for automotive applications
(2015)
We present a novel approach for improved hand-gesture recognition by a single time-of-flight(ToF) sensor in an automotive environment. As the sensor's lateral resolution is comparatively low, we employ a learning approach comprising multiple processing steps, including PCA-based cropping, the computation of robust point cloud descriptors and training of a Multilayer perceptron (MLP) on a large database of samples. A sophisticated temporal fusion technique boosts the overall robustness of recognition by taking into account data coming from previous classification steps. Overall results are very satisfactory when evaluated on a large benchmark set of ten different hand poses, especially when it comes to generalization on previously unknown persons.
We present a system for efficient dynamic hand gesture recognition based on a single time-of-flight sensor. As opposed to other approaches, we simply rely on depth data to interpret user movement with the hand in mid-air. We set up a large database to train multilayer perceptrons (MLPs) which are subsequently used for classification of static hand poses that define the targeted dynamic gestures. In order to remain robust against noise and to balance the low sensor resolution, PCA is used for data cropping and highly descriptive features, obtainable in real-time, are presented. Our simple yet efficient definition of a dynamic hand gesture shows how strong results are achievable in an automotive environment allowing for interesting and sophisticated applications to be realized.
We present a novel hierarchical approach to multi-class classification which is generic in that it can be applied to different classification models (e.g., support vector machines, perceptrons), and makes no explicit assumptions about the probabilistic structure of the problem as it is usually done in multi-class classification. By adding a cascade of additional classifiers, each of which receives the previous classifier's output in addition to regular input data, the approach harnesses unused information that manifests itself in the form of, e.g., correlations between predicted classes. Using multilayer perceptrons as a classification model, we demonstrate the validity of this approach by testing it on a complex ten-class 3D gesture recognition task.
Utilizing biometrie traits for privacy- and security-applications is receiving an increasing attention. Applications such as personal identification, access control, forensics appli-cations, e-banking, e-government, e-health and recently person-alized human-smart-home and human-robot interaction present some examples. In order to offer person-specific services for/of specific person a pre-identifying step should be done in the run-up. Using biometric in such application is encountered by diverse challenges. First, using one trait and excluding the others depends on the application aimed to. Some applications demand directly touch to biometric sensors, while others don't. Second challenge is the reliability of used biometric arrangement. Civilized application demands lower reliability comparing to the forensics ones. And third, for biometric system could only one trait be used (uni-modal systems) or multiple traits (Bi- or Multi-modal systems). The latter is applied, when systems with a relative high reliability are expected. The main aim of this paper is providing a comprehensive view about biometric and its application. The above mentioned challenges will be analyzed deeply. The suitability of each biometric sensor according to the aimed application will be deeply discussed. Detailed com-parison between uni-modal and Multi-modal biometric system will present which system where to be utilized. Privacy and security issues of biometric systems will be discussed too. Three scenarios of biometric application in home-environment, human-robot-interaction and e-health will be presented.
As smart homes are being more and more popular, the needs of finding assisting systems which interface between users and home environments are growing. Furthermore, for people living in such homes, elderly and disabled people in particular and others in general, it is totally important to develop devices, which can support and aid them in their ordinary daily life. We focused in this work on sustaining privacy issues of the user during a real interaction with the surrounding home environment. A smart person-specific assistant system for services in home environment is proposed. The role of this system is the assisting of persons by controlling home activities and guiding the adaption of Smart-Home-Human interface towards the needs of the considered person. At the same time the system sustains privacy issues of it’s interaction partner. As a special case of medical assisting the system is so implemented, that it provides for elderly or disabled people person-specific medical assistance . The system has the ability of identifying its interaction partner using some biometric features. According to the recognized ID the system, first, adopts towards the needs of recognized person. Second the system represents person-specific list of medicines either visually or auditive. And third the system gives an alarm in the case of taking medicament either later or earlier as normal taking time.
Forschung an Hochschulen
(2015)
In diesem Aufsatz soll die Forschung an Fachhochschulen beispielhaft aus dem Blickwinkel des Instituts Informatik der in 2009 gegründeten Hochschule Ruhr West betrachtet werden. Am Institut Informatik ist es das Ziel Lehre und Forschung geeignet zu verknüpfen, um Studierenden, wissenschaftlichen Mitarbeiterinnen und Mitarbeitern und auch Lehrenden ein attraktives Angebot in Forschung und Lehre im Bereich der Informatik zu liefern. Dabei bilden neben der Durchführung interessanter Lehrveranstaltungen, welche durch aktuelle Forschungsfragestellungen angereichert werden, das kooperative Bearbeiten von gesellschaftlich relevanten und zukunftsweisenden Forschungsaufgaben, die Teilnahme an Forschungsverbünden, bilaterale Forschungsaktivitäten mit Partnern aus der Wirtschaft und das Einwerben von externen Mitteln, die Basis der Arbeit am Institut.
This contribution presents a novel approach of utilizing Time-of-Flight (ToF) technology for mid-air hand gesture recognition on mobile devices. ToF sensors are capable of providing depth data at high frame rates independent of illumination making any kind of application possible for in- and outdoor situations. This comes at the cost of precision regarding depth measurements and comparatively low lateral resolution. We present a novel feature generation technique based on a rasterization of the point clouds which
realizes fixed-sized input making Deep Learning approaches applicable using Convolutional Neural Networks. In order to increase precision we introduce several methods to reduce noise and normalize the input to overcome difficulties in scaling. Backed by a large-scale database of about half
a million data samples taken from different individuals our
contribution shows how hand gesture recognition is realiz-
able on commodity tablets in real-time at frame rates of up to 17Hz. A leave-one out cross-validation experiment
demonstrates the feasibility of our approach with classification errors as low as 1,5% achieved persons unknown to the model.
We present a light-weight real-time applicable 3D-gesture recognition system on mobile devices for improved Human-Machine Interaction. We utilize time-of-flight data coming from a single sensor and implement the whole gesture recognition pipeline on two different devices outlining the potential of integrating these sensors onto mobile devices. The main components are responsible for cropping the data to the essentials, calculation of meaningful features, training and classifying via neural networks and realizing a GUI on the device. With our system we achieve recognition rates of up to 98% on a 10-gesture set with frame rates reaching 20Hz, more than sufficient for any real-time applications.
We present a publicly available benchmark database for the problem of hand posture recognition from noisy depth data and fused RGB-D data obtained from low-cost time-of-flight (ToF) sensors. The database is the most extensive database of this kind containing over a million data samples (point clouds) recorded from 35 different individuals for ten different static hand postures. This captures a great amount of variance, due to person-related factors, but also scaling, translation and rotation are explicitly represented. Benchmark results achieved with a standard classification algorithm are computed by cross-validation both over samples and persons, the latter implying training on all persons but one and testing on the remaining one. An important result using this database is that cross-validation performance over samples (which is the standard procedure in machine learning) is systematically higher than cross-validation performance over persons, which is to our mind the true application-relevant measure of generalization performance.
Touch versus mid-air gesture interfaces in road scenarios-measuring driver performance degradation
(2016)
We present a study aimed at comparing the degradation of the driver's performance during touch gesture vs mid-air gesture use for infotainment system control. To this end, 17 participants were asked to perform the Lane Change Test. This requires each participant to steer a vehicle in a simulated driving environment while interacting with an infotainment system via touch and mid-air gestures. The decrease in performance is measured as the deviation from an optimal baseline. This study concludes comparable deviations from the baseline for the secondary task of infotainment interaction for both interaction variants. This is significant as all participants are experienced in touch interaction, however have had no experience at all with mid-air gesture interaction, favoring mid-air gestures for the long-term scenario.
Given the success of convolutional neural networks (CNNs) during recent years in numerous object recognition tasks, it seems logical to further extend their applicability to the treatment of three-dimensional data such as point clouds provided by depth sensors. To this end, we present an approach exploiting the CNN’s ability of automated feature generation and combine it with a novel 3D feature computation technique, preserving local information contained in the data. Experiments are conducted on a large data set of 600.000 samples of hand postures obtained via ToF (time-of-flight) sensors from 20 different persons, after an extensive parameter search in order to optimize network structure. Generalization performance, measured by a leave-one-person-out scheme, exceeds that of any other method presented for this specific task, bringing the error for some persons down to 1.5 %.
Applying step heating thermography to wind turbine rotor blades as a non-destructive testing method
(2017)
Systems for automated image analysis are useful for a variety of tasks and their importance is still growing due to technological advances and an increase of social acceptance. Especially in the field of driver assistance systems the progress in science has reached a level of high performance. Fully or partly autonomously guided vehicles, particularly for road-based traffic, pose high demands on the development of reliable algorithms due to the conditions imposed by natural environments. At the Institut für Neuroinformatik methods for analyzing driving relevant scenes by computer vision are developed in cooperation with several partners from the automobile industry. We introduce a system which extracts the important information from an image taken by a CCD camera installed at the rear view mirror in a car. The approach consists of a sequential and a parallel sensor and information processing. Three main tasks namely the initial segmentation (object detection), the object tracking and the object classification are realized by integration in the sequential branch and by fusion in the parallel branch. The main gain of this approach is given by the integrative coupling of different algorithms providing partly redundant information.
Building upon prior results, we present an alternative approach to efficiently classifying a complex set of 3D hand poses obtained from modern Time-Of-Flight-Sensors (TOF). We demonstrate it is possible to achieve satisfactory results in spite of low resolution and high noise (inflicted by the sensors) and a demanding outdoor environment. We set up a large database of pointclouds in order to train multilayer perceptrons as well as support vector machines to classify the various hand poses. Our goal is to fuse data from multiple TOF sensors, which observe the poses from multiple angles. The presented contribution illustrates that real-time capability can be maintained with such a setup as the used 3D descriptors, the fusion strategy as well as the online confidence measures are computationally efficient.
In this article we present a system for coupling different base algorithms and sensors for segmentation. Three different solutions for image segmentation by fusion are described, compared and results are shown. The fusion of base algorithms with colorinformation and a sensor fusion process of an optical and a radar sensor including a feedback over time is realized. A feature-in decision-out fusion process is solved. For the fusion process a multi layer perceptron (MLP) with one hidden layer is used as a coupling net. The activity of the output neuron represents the membership of each pixel to an initial segment.
We present a novel approach of distributing small-to mid-scale neural networks onto modern parallel architectures. In this context we discuss the induced challenges and possible solutions. We provide a detailed theoretical analysis with respect to space and time complexities and reinforce our computation model with evaluations which show a performance gain over state of the art approaches.
Object detection systems which operate on large data streams require an efficient scaling with available computation power. We analyze how the use of tile-images can increase the efficiency (i.e. execution speed) of distributed HOG-based object detectors. Furthermore we discuss the challenges of using our developed algorithms in practical large scale scenarios. We show with a structured evaluation that our approach can provide a speed-up of 30-180 % for existing architectures. Due to the its generic formulation it can be applied to a wide range of HOG-based (or similar) algorithms. In this context we also study the effects of applying our method to an existing detector and discuss a scalable strategy for distributing the computation among nodes in a cluster system.
The behavior planning of a vehicle in real world traffic is a difficult problem to be solved. If different hierarchies of tasks and purposes are built to structure the behavior of a driver, complex systems can be designed. But finally behavior planning in vehicles can only influence the controlled variables: steering angle and velocity. In this paper a behavior planning for a driver assistance system aiming on cruise control is proposed. In this system the controlled variables are determined by an evaluation of the dynamics of two one-dimensional neural fields. The stimuli of the field are determined according to sensor information produced by a simulation environment.
In this paper, we describe a method to model human clothes for a later recognition by the use of RGB- and SWIR-cameras. A basic model is estimated during people detection and tracking. This model will be refined if the recognition is triggered. For the refining, several saliency maps are used to extract individual features. These individual features are located separately for any human body parts. The body parts are estimated by the use of a silhouette extraction combined with a skeleton estimation. In this way, the model describes the human clothes in a compact manner which allows the use of a simple and fast comparison method for people recognition. Such models can be used in security and service applications.
A self-driving car that operates on the SAE automation level 3 or 4 can navigate through different traffic conditions without human input. If such a system is on its operating limits, it will emit a takeover request before shutting down. This request will likely generate a physical response of the driver. Our goal is to shed light on the stress perception of drivers in various scenarios. To this end, we have carried out a feasibility study for preparation. Two subjects drove an autonomous vehicle and during the ride ECG signals were recorded, and afterwards evaluated. Unfortunately, the stress reaction to takeover requests could not be investigated, due to the poor function of the autonomous driving mode from the vehicle, however the reaction to autopilot misconduct without warning to the driver could be investigated instead.
Checking wind turbines for damage is a common problem for operators of wind parks, as regular inspections are legally required in many countries and prevention is economically viable. While some of the common forms of damage are easily visible on the surface, structural problems can remain invisible for years before they eventually result in catastrophic failure of a rotor blade. Common forms of testing fibre composite parts like ultrasonic testing or X-ray tests are impractical due to the large dimensions of wind turbine components and their limited accessibility for any short-range methods. Active thermographic inspection of wind turbines is a promising approach to testing for structural flaws beneath the surface of rotor blades. As part of an ongoing research project, a setup for testing the general viability of this method was built and used to compare different thermographic cameras. A sample cut from a discarded rotor blade was modified to emulate structural damage. The results are promising for the development of a cost effective on-site testing system.
Increasing economic viability and safety through structural health monitoring of wind turbines
(2017)
Serious accidents with property damage or even human casualties, result from structural flaws in wind turbine rotor blades. Common maintenance practices result in long downtimes and do not lead to the required results. Therefore, the Ruhr West University of Applied Sciences and the iQbis Consulting GmbH, currently research a new structural health monitoring method for wind turbine rotor blades. The goal of this project is to build a sensor system that can detect structural weaknesses inside of rotor blades without the need of downtime for industrial climbers. This technology has the potential to prevent accidents, save lives, extend the useful life of wind turbines and optimize the production of green energy.
We present a pipeline for recognizing dynamic freehand gestures on mobile devices based on extracting depth information coming from a single Time-of-Flight sensor. Hand gestures are recorded with a mobile 3D sensor, transformed frame by frame into an appropriate 3D descriptor and fed into a deep LSTM network for recognition purposes. LSTM being a recurrent neural model, it is uniquely suited for classifying explicitly time-dependent data such as hand gestures. For training and testing purposes, we create a small database of four hand gesture classes, each comprising 40 × 150 3D frames. We conduct experiments concerning execution speed on a mobile device, generalization capability as a function of network topology, and classification ability ‘ahead of time’, i.e., when the gesture is not yet completed. Recognition rates are high (>95%) and maintainable in real-time as a single classification step requires less than 1 ms computation time, introducing freehand gestures for mobile systems.
RELEVANCE & RESEARCH QUESTION: Currently the effectiveness of Virtual Reality (VR) and Augmented Reality (AR) systems as practice teaching methods are virtually uncharted. The proof that these systems can provide the same or better learning outcomes than a text instructed practical task could represent a significant benefit for educational activities. METHODS & DATA: To fathom the effectiveness, an experimental study with the three conditions (VR, AR and a real setup) were used to teach participant how to assemble a standard computer. Each condition was divided into two parts: part one in which participants were confronted with their specific scenario, part two in which participants had to go through a real practice after one week. The learning outcome was determined by the designation of hardware parts, a quiz that queried their function and the correct assembling of the components in addition to needed time. Apart from the mere performance, the acceptance of such application in academic context and difference in evaluation by men and women were of interest. RESULTS: Results concerning the Learning Outcome showed that participants from the VR condition outperformed those learned from the real setup ((M=10.0, SD=0.0) [virtual reality] vs. (M=8.95, SD=1.27) [control]). Furthermore, results from the assembling duration assessment demonstrated that VR Group Participants completed their tasks 6.62% faster than the control group. Regarding the identification of Hardware Parts, both groups scored a significant improvement during the post condition compared to the first test run, indicating a learning progress. However, due to the VR group achieving a better outcome in average answers and a more significant difference between the trials, the results indicate a better performance by participants assigned to the VR condition. ADDED VALUE: The results revealed that VR and AR systems could exceed text-based approach in terms of learning outcome performance. The effectiveness of the systems implicates a major benefit for the educational landscape, as learning content that is not realizable in terms of cost, distance or logistics could be designed as an immersive and engaging experience.
Artificial Intelligence Driven Human-Machine Collaboration Scenarios in Virtual Reality (Poster)
(2018)
In this review, we describe current Machine Learning approaches to hand gesture recognition with depth data from time-of-flight sensors. In particular, we summarise the achievements on a line of research at the Computational Neuroscience laboratory at the Ruhr West University of Applied Sciences. Relating our results to the work of others in this field, we confirm that Convolutional Neural Networks and Long Short-Term Memory yield most reliable results. We investigated several sensor data fusion techniques in a deep learning framework and performed user studies to evaluate our system in practice. During our course of research, we gathered and published our data in a novel benchmark dataset (REHAP), containing over a million unique three-dimensional hand posture samples.
Relax yourself - Using Virtual Reality to enhance employees mental health and work performance
(2019)
This paper presents work-in-progress aiming to develop an actively adapting virtual reality (VR) relaxation application. Due to the immersive nature of VR technologies, people can escape from their real environment and get into a relaxing state. Goal of the application is to adapt to the users' physiological signals to foster the positive effect. Until now, a first version of the VR application was constructed and is currently evaluated in an experiment. Preliminary results of this study demonstrate that people appreciate the immersion into the virtual environment and escape from reality. Moreover, participants highlighted the option to adapt users' needs and preferences. Based on the final study data, the constructed application will be enhanced with regard to adoption and surrounding factors.
Die Entwicklung von vollautomatisierten Fahrzeugen wird in der gesellschaftlichen Diskussion immer präsenter. Wichtig für die Durchsetzung und verbreitete Nutzung dieser technischer Neuerungen ist jedoch vor allem die Akzeptanz der Bevölkerung – in diesem Fall nicht nur die der potenziellen KäuferInnen sondern auch die der übrigen Verkehrs-teilnehmenden. Vorgestellt wird eine explorative Online-Studie zur Akzeptanz von auto-nomen Fahren basierend auf quantitativen und qualitativen Daten einer Stichprobe von N = 89. Die Ergebnisse zeigen unter anderem eine geringe Vertrautheit mit dem Thema, ein vergleichsweise ausgeprägtes Vertrauen aber eine geringe Nutzungsabsicht.
Mobile Walzenmesstechnik
(2003)
