Refine
H-BRS Bibliography
- yes (970) (remove)
Departments, institutes and facilities
- Fachbereich Informatik (970) (remove)
Document Type
- Conference Object (574)
- Article (186)
- Preprint (45)
- Part of a Book (42)
- Report (33)
- Book (monograph, edited volume) (31)
- Doctoral Thesis (22)
- Conference Proceedings (17)
- Research Data (11)
- Contribution to a Periodical (5)
Year of publication
Has Fulltext
- no (970) (remove)
Keywords
- Robotics (12)
- Virtual Reality (10)
- virtual reality (8)
- Quality diversity (7)
- 3D user interface (6)
- Lehrbuch (6)
- Machine Learning (6)
- Augmented Reality (5)
- Benchmarking (5)
- CUDA (5)
In den letzten Jahren haben sich elektronische Zahlungssysteme als populäre Alternative zur klassischen Bargeldzahlung etabliert. Diese Zahlungssysteme bestehen in der Regel aus zwei elementaren Komponenten: einem Terminal und einer Kasse. Damit ist der Käufer eines Produktes in der Lage, seine Schuld gegenüber dem Verkäufer bargeldlos und elektronisch zu begleichen. Die dabei am Häufigsten anfallenden Geschäftsprozesse, das Buchen und das Stornieren von Zahlungsbelegen, werden hierbei als Transaktionen bezeichnet, da diese entweder vollständig gelingen oder im Fehlerfall ohne Auswirkungen bleiben müssen. In diesem Buch wird daher die Implementierung eines zuverlässigen Zahlungssystems mit einem TeleCash-Terminal dargestellt. Dabei werden in den geforderten Geschäftsprozessen die wichtigen Transaktionseigenschaften sichergestellt. Es werden dazu zunächst die Grundlagen von Transaktionen erarbeitet und ein geeignetes Transaktionskonzept entwickelt. Anschließend wird die konkrete Realisierung des Systems mit Hilfe der Java Transaction Services durchgeführt. Abschließend wird das entstandene System hinsichtlich seiner Transaktionseigenschaften untersucht.
Current robot platforms are being employed to collaborate with humans in a wide range of domestic and industrial tasks. These environments require autonomous systems that are able to classify and communicate anomalous situations such as fires, injured persons, car accidents; or generally, any potentially dangerous situation for humans. In this paper we introduce an anomaly detection dataset for the purpose of robot applications as well as the design and implementation of a deep learning architecture that classifies and describes dangerous situations using only a single image as input. We report a classification accuracy of 97 % and METEOR score of 16.2. We will make the dataset publicly available after this paper is accepted.
In this paper we introduce the Perception for Autonomous Systems (PAZ) software library. PAZ is a hierarchical perception library that allow users to manipulate multiple levels of abstraction in accordance to their requirements or skill level. More specifically, PAZ is divided into three hierarchical levels which we refer to as pipelines, processors, and backends. These abstractions allows users to compose functions in a hierarchical modular scheme that can be applied for preprocessing, data-augmentation, prediction and postprocessing of inputs and outputs of machine learning (ML) models. PAZ uses these abstractions to build reusable training and prediction pipelines for multiple robot perception tasks such as: 2D keypoint estimation, 2D object detection, 3D keypoint discovery, 6D pose estimation, emotion classification, face recognition, instance segmentation, and attention mechanisms.
Emotion and gender recognition from facial features are important properties of human empathy. Robots should also have these capabilities. For this purpose we have designed special convolutional modules that allow a model to recognize emotions and gender with a considerable lower number of parameters, enabling real-time evaluation on a constrained platform. We report accuracies of 96% in the IMDB gender dataset and 66% in the FER-2013 emotion dataset, while requiring a computation time of less than 0.008 seconds on a Core i7 CPU. All our code, demos and pre-trained architectures have been released under an open-source license in our repository at https://github.com/oarriaga/face classification.
In this paper we propose an implement a general convolutional neural network (CNN) building framework for designing real-time CNNs. We validate our models by creating a real-time vision system which accomplishes the tasks of face detection, gender classification and emotion classification simultaneously in one blended step using our proposed CNN architecture. After presenting the details of the training procedure setup we proceed to evaluate on standard benchmark sets. We report accuracies of 96% in the IMDB gender dataset and 66% in the FER-2013 emotion dataset. Along with this we also introduced the very recent real-time enabled guided back-propagation visualization technique. Guided back-propagation uncovers the dynamics of the weight changes and evaluates the learned features. We argue that the careful implementation of modern CNN architectures, the use of the current regularization methods and the visualization of previously hidden features are necessary in order to reduce the gap between slow performances and real-time architectures. Our system has been validated by its deployment on a Care-O-bot 3 robot used during RoboCup@Home competitions. All our code, demos and pre-trained architectures have been released under an open-source license in our public repository.
The goal of this work is to develop an integration framework for a robotic software system which enables robotic learning by experimentation within a distributed and heterogeneous setting. To meet this challenge, the authors specified, defined, developed, implemented and tested a component-based architecture called XPERSIF. The architecture comprises loosely-coupled, autonomous components that offer services through their well-defined interfaces and form a service-oriented architecture. The Ice middleware is used in the communication layer. Additionally, the successful integration of the XPERSim simulator into the system has enabled simultaneous quasi-realtime observation of the simulation by numerous, distributed users.
Adapting plans to changes in the environment by finding alternatives and taking advantage of opportunities is a common human behavior. The need for such behavior is often rooted in the uncertainty produced by our incomplete knowledge of the environment. While several existing planning approaches deal with such issues, artificial agents still lack the robustness that humans display in accomplishing their tasks. In this work, we address this brittleness by combining Hierarchical Task Network planning, Description Logics, and the notions of affordances and conceptual similarity. The approach allows a domestic service robot to find ways to get a job done by making substitutions. We show how knowledge is modeled, how the reasoning process is used to create a constrained planning problem, and how the system handles cases where plan generation fails due to missing/unavailable objects. The results of the evaluation for two tasks in a domestic service domain show the viability of the approach in finding and making the appropriate goal transformations.
Humans exhibit flexible and robust behavior in achieving their goals. We make suitable substitutions for objects, actions, or tools to get the job done. When opportunities that would allow us to reach our goals with less effort arise, we often take advantage of them. Robots are not nearly as robust in handling such situations. Enabling a domestic service robot to find ways to get a job done by making substitutions is the goal of our work. In this paper, we highlight the challenges faced in our approach to combine Hierarchical Task Network planning, Description Logics, and the notions of affordances and conceptual similarity. We present open questions in modeling the necessary knowledge, creating planning problems, and enabling the system to handle cases where plan generation fails due to missing/unavailable objects.
In this paper, we present XPERSim, a 3D simulator built on top of open source components that enables users to quickly and easily construct an accurate and photo-realistic simulation for robots of arbitrary morphology and their environments. While many existing robot simulators provide a good dynamics simulation, they often lack the high quality visualization that is now possible with general-purpose hardware. XPERSim achieves such visualization by using the Object-Oriented Graphics Rendering Engine 3D (Ogre) engine to render the simulation whose dynamics are calculated using the Open Dynamics Engine (ODE). Through XPERSim’s integration into a component-based software integration framework used for robotic learning by experimentation, XPERSIF, and the use of the scene-oriented nature of the Ogre engine, the simulation is distributed to numerous users that include researchers and robotic components, thus enabling simultaneous, quasi-realtime observation of the multiple-camera simulations.
Representing 3D surfaces as level sets of continuous functions over R3 is the common denominator of neural implicit representations, which recently enabled remarkable progress in geometric deep learning and computer vision tasks. In order to represent 3D motion within this framework, it is often assumed (either explicitly or implicitly) that the transformations which a surface may undergo are homeomorphic: this is not necessarily true, for instance, in the case of fluid dynamics. In order to represent more general classes of deformations, we propose to apply this theoretical framework as regularizers for the optimization of simple 4D implicit functions (such as signed distance fields). We show that our representation is capable of capturing both homeomorphic and topology-changing deformations, while also defining correspondences over the continuously-reconstructed surfaces.
The majority of biomedical knowledge is stored in structured databases or as unstructured text in scientific publications. This vast amount of information has led to numerous machine learning-based biological applications using either text through natural language processing (NLP) or structured data through knowledge graph embedding models (KGEMs). However, representations based on a single modality are inherently limited. To generate better representations of biological knowledge, we propose STonKGs, a Sophisticated Transformer trained on biomedical text and Knowledge Graphs. This multimodal Transformer uses combined input sequences of structured information from KGs and unstructured text data from biomedical literature to learn joint representations. First, we pre-trained STonKGs on a knowledge base assembled by the Integrated Network and Dynamical Reasoning Assembler (INDRA) consisting of millions of text-triple pairs extracted from biomedical literature by multiple NLP systems. Then, we benchmarked STonKGs against two baseline models trained on either one of the modalities (i.e., text or KG) across eight different classification tasks, each corresponding to a different biological application. Our results demonstrate that STonKGs outperforms both baselines, especially on the more challenging tasks with respect to the number of classes, improving upon the F1-score of the best baseline by up to 0.083. Additionally, our pre-trained model as well as the model architecture can be adapted to various other transfer learning applications. Finally, the source code and pre-trained STonKGs models are available at https://github.com/stonkgs/stonkgs and https://huggingface.co/stonkgs/stonkgs-150k.
In this contribution a machine vision inspection system is presented which is designed as a length measuring sensor. It is developed to be applied to a range of heat shrink tubes, varying in length, diameter and color. The challenges of this task were the precision and accuracy demands as well as the real-time applicability of the entire approach since it should be realized in regular industrial line production. In production, heat shrink tubes are cut to specific sizes from a continuous tube. A multi-measurement strategy has been developed, which measures each individual tube segment several times with sub pixel accuracy while being in the visual field. The developed approach allows for a contact-free and fully automatic control of 100% of produced heat shrink tubes according to the given requirements with a measuring precision of 0.1mm. Depending on the color, length and diameter of the tubes considered, a true positive rate of 99.99% to 100% has been reached at a true negative rate of > 99.7.
Diese Arbeit beschäftigt sich mit der Effizienz der Seitenkanal-Kryptanalyse. In Teil II dieser Arbeit demonstrieren wir, wie die Laufzeit der wichtigsten Analysewerkzeuge mit Hilfe der CUDA Plattform erheblich gesteigert werden kann. Zweitens untersuchen wir neue Ansätze der profilierenden Seitenkanal-Kryptanalyse. Der Forschungszweig des maschinellen Lernens kann für deutliche Verbesserungen adaptiert werden, wurde jedoch wenig dahingehend untersucht. In Teil III dieser Arbeit präsentieren wir zwei neue Methoden, die einige Gemeinsamkeiten jedoch auch einige Unterschiede aufbieten, sodass sich Prüfergebnisse in einem vollständigeren Bild zeigen lassen. Darüber hinaus schlagen wir in Teil IV eine Seitenkanalanwendung zum Schutz geistigen Eigentums (IP) vor. In Teil V beschäftigen wir uns tiefergehend mit praktischer Seitenkanal-Kryptanalyse, indem wir Attacken auf einen Sicherheitsmikrokontroller durchführen, der Anwendung in einer, in Deutschland weit verbreiteten, EC Karte findet.