Refine
H-BRS Bibliography
- yes (69) (remove)
Departments, institutes and facilities
- Fachbereich Informatik (69) (remove)
Document Type
- Conference Object (37)
- Article (18)
- Doctoral Thesis (3)
- Preprint (3)
- Report (3)
- Part of a Book (2)
- Book (monograph, edited volume) (1)
- Research Data (1)
- Master's Thesis (1)
Year of publication
- 2019 (69) (remove)
Keywords
- Navigation (3)
- Drosophila (2)
- Hyperspectral image (2)
- Raman microscopy (2)
- Ray tracing (2)
- UAV (2)
- Virtual Reality (2)
- aerodynamics (2)
- dynamic vector fields (2)
- flight zone (2)
Traffic sign recognition is an important component of many advanced driving assistance systems, and it is required for full autonomous driving. Computational performance is usually the bottleneck in using large scale neural networks for this purpose. SqueezeNet is a good candidate for efficient image classification of traffic signs, but in our experiments it does not reach high accuracy, and we believe this is due to lack of data, requiring data augmentation. Generative adversarial networks can learn the high dimensional distribution of empirical data, allowing the generation of new data points. In this paper we apply pix2pix GANs architecture to generate new traffic sign images and evaluate the use of these images in data augmentation. We were motivated to use pix2pix to translate symbolic sign images to real ones due to the mode collapse in Conditional GANs. Through our experiments we found that data augmentation using GAN can increase classification accuracy for circular traffic signs from 92.1% to 94.0%, and for triangular traffic signs from 93.8% to 95.3%, producing an overall improvement of 2%. However some traditional augmentation techniques can outperform GAN data augmentation, for example contrast variation in circular traffic signs (95.5%) and displacement on triangular traffic signs (96.7 %). Our negative results shows that while GANs can be naively used for data augmentation, they are not always the best choice, depending on the problem and variability in the data.
Background: Virtual reality combined with spherical treadmills is used across species for studying neural circuits underlying navigation.
New Method: We developed an optical flow-based method for tracking treadmil ball motion in real-time using a single high-resolution camera.
Results: Tracking accuracy and timing were determined using calibration data. Ball tracking was performed at 500 Hz and integrated with an open source game engine for virtual reality projection. The projection was updated at 120 Hz with a latency with respect to ball motion of 30 ± 8 ms.
Comparison: with Existing Method(s) Optical flow based tracking of treadmill motion is typically achieved using optical mice. The camera-based optical flow tracking system developed here is based on off-the-shelf components and offers control over the image acquisition and processing parameters. This results in flexibility with respect to tracking conditions – such as ball surface texture, lighting conditions, or ball size – as well as camera alignment and calibration.
Conclusions: A fast system for rotational ball motion tracking suitable for virtual reality animal behavior across different scales was developed and characterized.
Data-Driven Robot Fault Detection and Diagnosis Using Generative Models: A Modified SFDD Algorithm
(2019)
This paper presents a modification of the data-driven sensor-based fault detection and diagnosis (SFDD) algorithm for online robot monitoring. Our version of the algorithm uses a collection of generative models, in particular restricted Boltzmann machines, each of which represents the distribution of sliding window correlations between a pair of correlated measurements. We use such models in a residual generation scheme, where high residuals generate conflict sets that are then used in a subsequent diagnosis step. As a proof of concept, the framework is evaluated on a mobile logistics robot for the problem of recognising disconnected wheels, such that the evaluation demonstrates the feasibility of the framework (on the faulty data set, the models obtained 88.6% precision and 75.6% recall rates), but also shows that the monitoring results are influenced by the choice of distribution model and the model parameters as a whole.
Tell Your Robot What To Do: Evaluation of Natural Language Models for Robot Command Processing
(2019)
The use of natural language to indicate robot tasks is a convenient way to command robots. As a result, several models and approaches capable of understanding robot commands have been developed, which however complicates the choice of a suitable model for a given scenario. In this work, we present a comparative analysis and benchmarking of four natural language understanding models - Mbot, Rasa, LU4R, and ECG. We particularly evaluate the performance of the models to understand domestic service robot commands by recognizing the actions and any complementary information in them in three use cases: the RoboCup@Home General Purpose Service Robot (GPSR) category 1 contest, GPSR category 2, and hospital logistics in the context of the ROPOD project.
In Sensor-based Fault Detection and Diagnosis (SFDD) methods, spatial and temporal dependencies among the sensor signals can be modeled to detect faults in the sensors, if the defined dependencies change over time. In this work, we model Granger causal relationships between pairs of sensor data streams to detect changes in their dependencies. We compare the method on simulated signals with the Pearson correlation, and show that the method elegantly handles noise and lags in the signals and provides appreciable dependency detection. We further evaluate the method using sensor data from a mobile robot by injecting both internal and external faults during operation of the robot. The results show that the method is able to detect changes in the system when faults are injected, but is also prone to detecting false positives. This suggests that this method can be used as a weak detection of faults, but other methods, such as the use of a structural model, are required to reliably detect and diagnose faults.
For robots acting - and failing - in everyday environments, a predictable behaviour representation is important so that it can be utilised for failure analysis, recovery, and subsequent improvement. Learning from demonstration combined with dynamic motion primitives is one commonly used technique for creating models that are easy to analyse and interpret; however, mobile manipulators complicate such models since they need the ability to synchronise arm and base motions for performing purposeful tasks. In this paper, we analyse dynamic motion primitives in the context of a mobile manipulator - a Toyota Human Support Robot (HSR)- and introduce a small extension of dynamic motion primitives that makes it possible to perform whole body motion with a mobile manipulator. We then present an extensive set of experiments in which our robot was grasping various everyday objects in a domestic environment, where a sequence of object detection, pose estimation, and manipulation was required for successfully completing the task. Our experiments demonstrate the feasibility of the proposed whole body motion framework for everyday object manipulation, but also illustrate the necessity for highly adaptive manipulation strategies that make better use of a robot's perceptual capabilities.
PosturePairsDB19
(2019)
The application of Raman and infrared (IR) microspectroscopy is leading to hyperspectral data containing complementary information concerning the molecular composition of a sample. The classification of hyperspectral data from the individual spectroscopic approaches is already state-of-the-art in several fields of research. However, more complex structured samples and difficult measuring conditions might affect the accuracy of classification results negatively and could make a successful classification of the sample components challenging. This contribution presents a comprehensive comparison in supervised pixel classification of hyperspectral microscopic images, proving that a combined approach of Raman and IR microspectroscopy has a high potential to improve classification rates by a meaningful extension of the feature space. It shows that the complementary information in spatially co-registered hyperspectral images of polymer samples can be accessed using different feature extraction methods and, once fused on the feature-level, is in general more accurately classifiable in a pattern recognition task than the corresponding classification results for data derived from the individual spectroscopic approaches.
Herein we report an update to ACPYPE, a Python3 tool that now properly converts AMBER to GROMACS topologies for force fields that utilize nondefault and nonuniform 1–4 electrostatic and nonbonded scaling factors or negative dihedral force constants. Prior to this work, ACPYPE only converted AMBER topologies that used uniform, default 1–4 scaling factors and positive dihedral force constants. We demonstrate that the updated ACPYPE accurately transfers the GLYCAM06 force field from AMBER to GROMACS topology files, which employs non-uniform 1–4 scaling factors as well as negative dihedral force constants. Validation was performed using β-d-GlcNAc through gas-phase analysis of dihedral energy curves and probability density functions. The updated ACPYPE retains all of its original functionality, but now allows the simulation of complex glycomolecular systems in GROMACS using AMBER-originated force fields. ACPYPE is available for download at https://github.com/alanwilter/acpype.
Computer graphics research strives to synthesize images of a high visual realism that are indistinguishable from real visual experiences. While modern image synthesis approaches enable to create digital images of astonishing complexity and beauty, processing resources remain a limiting factor. Here, rendering efficiency is a central challenge involving a trade-off between visual fidelity and interactivity. For that reason, there is still a fundamental difference between the perception of the physical world and computer-generated imagery. At the same time, advances in display technologies drive the development of novel display devices. The dynamic range, the pixel densities, and refresh rates are constantly increasing. Display systems enable a larger visual field to be addressed by covering a wider field-of-view, due to either their size or in the form of head-mounted devices. Currently, research prototypes are ranging from stereo and multi-view systems, head-mounted devices with adaptable lenses, up to retinal projection, and lightfield/holographic displays. Computer graphics has to keep step with, as driving these devices presents us with immense challenges, most of which are currently unsolved. Fortunately, the human visual system has certain limitations, which means that providing the highest possible visual quality is not always necessary. Visual input passes through the eye’s optics, is filtered, and is processed at higher level structures in the brain. Knowledge of these processes helps to design novel rendering approaches that allow the creation of images at a higher quality and within a reduced time-frame. This thesis presents the state-of-the-art research and models that exploit the limitations of perception in order to increase visual quality but also to reduce workload alike - a concept we call perception-driven rendering. This research results in several practical rendering approaches that allow some of the fundamental challenges of computer graphics to be tackled. By using different tracking hardware, display systems, and head-mounted devices, we show the potential of each of the presented systems. The capturing of specific processes of the human visual system can be improved by combining multiple measurements using machine learning techniques. Different sampling, filtering, and reconstruction techniques aid the visual quality of the synthesized images. An in-depth evaluation of the presented systems including benchmarks, comparative examination with image metrics as well as user studies and experiments demonstrated that the methods introduced are visually superior or on the same qualitative level as ground truth, whilst having a significantly reduced computational complexity.
Treatment options for acute myeloid leukemia (AML) remain extremely limited and associated with significant toxicity. Nicotinamide phosphoribosyltransferase (NAMPT) is involved in the generation of NAD+ and a potential therapeutic target in AML. We evaluated the effect of KPT-9274, a p21-activated kinase 4/NAMPT inhibitor that possesses a unique NAMPT-binding profile based on in silico modeling compared with earlier compounds pursued against this target. KPT-9274 elicited loss of mitochondrial respiration and glycolysis and induced apoptosis in AML subtypes independent of mutations and genomic abnormalities. These actions occurred mainly through the depletion of NAD+, whereas genetic knockdown of p21-activated kinase 4 did not induce cytotoxicity in AML cell lines or influence the cytotoxic effect of KPT-9274. KPT-9274 exposure reduced colony formation, increased blast differentiation, and diminished the frequency of leukemia-initiating cells from primary AML samples; KPT-9274 was minimally cytotoxic toward normal hematopoietic or immune cells. In addition, KPT-9274 improved overall survival in vivo in 2 different mouse models of AML and reduced tumor development in a patient-derived xenograft model of AML. Overall, KPT-9274 exhibited broad preclinical activity across a variety of AML subtypes and warrants further investigation as a potential therapeutic agent for AML.
In an effort to assist researchers in choosing basis sets for quantum mechanical modeling of molecules (i.e. balancing calculation cost versus desired accuracy), we present a systematic study on the accuracy of computed conformational relative energies and their geometries in comparison to MP2/CBS and MP2/AV5Z data, respectively. In order to do so, we introduce a new nomenclature to unambiguously indicate how a CBS extrapolation was computed. Nineteen minima and transition states of buta-1,3-diene, propan-2-ol and the water dimer were optimized using forty-five different basis sets. Specifically, this includes one Pople (i.e. 6-31G(d)), eight Dunning (i.e. VXZ and AVXZ, X=2-5), twenty-five Jensen (i.e. pc-n, pcseg-n, aug-pcseg-n, pcSseg-n and aug-pcSseg-n, n=0-4) and nine Karlsruhe (e.g. def2-SV(P), def2-QZVPPD) basis sets. The molecules were chosen to represent both common and electronically diverse molecular systems. In comparison to MP2/CBS relative energies computed using the largest Jensen basis sets (i.e. n=2,3,4), the use of smaller sizes (n=0,1,2 and n=1,2,3) provides results that are within 0.11--0.24 and 0.09-0.16 kcal/mol. To practically guide researchers in their basis set choice, an equation is introduced that ranks basis sets based on a user-defined balance between their accuracy and calculation cost. Furthermore, we explain why the aug-pcseg-2, def2-TZVPPD and def2-TZVP basis sets are very suitable choices to balance speed and accuracy.
Lower back pain is one of the most prevalent diseases in Western societies. A large percentage of European and American populations suffer from back pain at some point in their lives. One successful approach to address lower back pain is postural training, which can be supported by wearable devices, providing real-time feedback about the user’s posture. In this work, we analyze the changes in posture induced by postural training. To this end, we compare snapshots before and after training, as measured by the Gokhale SpineTracker™. Considering pairs of before and after snapshots in different positions (standing, sitting, and bending), we introduce a feature space, that allows for unsupervised clustering. We show that resulting clusters represent certain groups of postural changes, which are meaningful to professional posture trainers.
When developing robot functionalities, finite state machines are commonly used due to their straightforward semantics and simple implementation. State machines are also a natural implementation choice when designing robot experiments, as they generally lead to reproducible program execution. In practice, the implementation of state machines can lead to significant code repetition and may necessitate unnecessary code interaction when reparameterisation is required. In this paper, we present a small Python library that allows state machines to be specified, configured, and dynamically created using a minimal domain-specific language. We illustrate the use of the library in three different use cases - scenario definition in the context of the RoboCup@Home competition, experiment design in the context of the ROPOD project, as well as specification transfer between robots.
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.
Interactive Object Detection
(2019)
The success of state-of-the-art object detection methods depend heavily on the availability of a large amount of annotated image data. The raw image data available from various sources are abundant but non-annotated. Annotating image data is often costly, time-consuming or needs expert help. In this work, a new paradigm of learning called Active Learning is explored which uses user interaction to obtain annotations for a subset of the dataset. The goal of active learning is to achieve superior object detection performance with images that are annotated on demand. To realize active learning method, the trade-off between the effort to annotate (annotation cost) unlabeled data and the performance of object detection model is minimised.
Random Forests based method called Hough Forest is chosen as the object detection model and the annotation cost is calculated as the predicted false positive and false negative rate. The framework is successfully evaluated on two Computer Vision benchmark and two Carl Zeiss custom datasets. Also, an evaluation of RGB, HoG and Deep features for the task is presented.
Experimental results show that using Deep features with Hough Forest achieves the maximum performance. By employing Active Learning, it is demonstrated that performance comparable to the fully supervised setting can be achieved by annotating just 2.5% of the images. To this end, an annotation tool is developed for user interaction during Active Learning.
In mathematical modeling by means of performance models, the Fitness-Fatigue Model (FF-Model) is a common approach in sport and exercise science to study the training performance relationship. The FF-Model uses an initial basic level of performance and two antagonistic terms (for fitness and fatigue). By model calibration, parameters are adapted to the subject’s individual physical response to training load. Although the simulation of the recorded training data in most cases shows useful results when the model is calibrated and all parameters are adjusted, this method has two major difficulties. First, a fitted value as basic performance will usually be too high. Second, without modification, the model cannot be simply used for prediction. By rewriting the FF-Model such that effects of former training history can be analyzed separately – we call those terms preload – it is possible to close the gap between a more realistic initial performance level and an athlete's actual performance level without distorting other model parameters and increase model accuracy substantially. Fitting error of the preload-extended FF-Model is less than 32% compared to the error of the FF-Model without preloads. Prediction error of the preload-extended FF-Model is around 54% of the error of the FF-Model without preloads.
Survival of patients with pediatric acute lymphoblastic leukemia (ALL) after allogeneic hematopoietic stem cell transplantation (allo-SCT) is mainly compromised by leukemia relapse, carrying dismal prognosis. As novel individualized therapeutic approaches are urgently needed, we performed whole-exome sequencing of leukemic blasts of 10 children with post–allo-SCT relapses with the aim of thoroughly characterizing the mutational landscape and identifying druggable mutations. We found that post–allo-SCT ALL relapses display highly diverse and mostly patient-individual genetic lesions. Moreover, mutational cluster analysis showed substantial clonal dynamics during leukemia progression from initial diagnosis to relapse after allo-SCT. Only very few alterations stayed constant over time. This dynamic clonality was exemplified by the detection of thiopurine resistance-mediating mutations in the nucleotidase NT5C2 in 3 patients’ first relapses, which disappeared in the post–allo-SCT relapses on relief of selective pressure of maintenance chemotherapy. Moreover, we identified TP53 mutations in 4 of 10 patients after allo-SCT, reflecting acquired chemoresistance associated with selective pressure of prior antineoplastic treatment. Finally, in 9 of 10 children’s post–allo-SCT relapse, we found alterations in genes for which targeted therapies with novel agents are readily available. We could show efficient targeting of leukemic blasts by APR-246 in 2 patients carrying TP53 mutations. Our findings shed light on the genetic basis of post–allo-SCT relapse and may pave the way for unraveling novel therapeutic strategies in this challenging situation.
Application developers constitute an important part of a digital platform’s ecosystem. Knowledge about psychological processes that drive developer behavior in platform ecosystems is scarce. We build on the lead userness construct which comprises two dimensions, trend leadership and high expected benefits from a solution, to explain how developers’ innovative work behavior (IWB) is stimulated. We employ an efficiencyoriented and a social-political perspective to investigate the relationship between lead userness and IWB. The efficiency-oriented view resonates well with the expected benefit dimension of lead userness, while the social-political view might be interpreted as a reflection of trend leadership. Using structural equation modeling, we test our model with a sample of over 400 developers from three platform ecosystems. We find that lead userness is indirectly associated with IWB and the performance-enhancing view to be the stronger predictor of IWB. Finally, we unravel differences between paid and unpaid app developers in platform ecosystems.
Modern Monte-Carlo-based rendering systems still suffer from the computational complexity involved in the generation of noise-free images, making it challenging to synthesize interactive previews. We present a framework suited for rendering such previews ofstatic scenes using a caching technique that builds upon a linkless octree. Our approach allows for memory-efficient storage and constant-time lookup to cache diffuse illumination at multiple hitpoints along the traced paths. Non-diffuse surfaces are dealt with in a hybrid way in order to reconstruct view-dependent illumination while maintaining interactive frame rates. By evaluating the visual fidelity against ground truth sequences and by benchmarking, we show that our approach compares well to low-noise path traced results, but with a greatly reduced computational complexity allowing for interactive frame rates. This way, our caching technique provides a useful tool for global illumination previews and multi-view rendering.
Energy Profiles of the Ring Puckering of Cyclopentane, Methylcyclopentane and Ethylcyclopentane
(2019)
Quantifying Interference in WiLD Networks using Topography Data and Realistic Antenna Patterns
(2019)
Avoiding possible interference is a key aspect to maximize the performance in Wi-Fi based Long Distance networks. In this paper we quantify self-induced interference based on data derived from our testbed and match the findings against simulations. By enhancing current simulation models with two key elements we significantly reduce the deviation between testbed and simulation: the usage of detailed antenna patterns compared to the cone model and propagation modeling enhanced by license-free topography data. Based on the gathered data we discuss several possible optimization approaches such as physical separation of local radios, tuning the sensitivity of the transmitter and using centralized compared to distributed channel assignment algorithms. While our testbed is based on 5 GHz Wi-Fi, we briefly discuss the possible impact of our results to other frequency bands.
Bond graph software can simulate bond graph models without the user needing to manually derive equations. This offers the power to model larger and more complex systems than in the past. Multibond graphs (those with vector bonds) offer a compact model which further eases handling multibody systems. Although multibond graphs can be simulated successfully, the use of vector bonds can present difficulties. In addition, most qualitative, bond graph–based exploitation relies on the use of scalar bonds. This article discusses the main methods for simulating bond graphs of multibody systems, using a graphical software platform. The transformation between models with vector and scalar bonds is presented. The methods are then compared with respect to both time and accuracy, through simulation of two benchmark models. This article is a tutorial on the existing methods for simulating three-dimensional rigid and holonomic multibody systems using bond graphs and discusses the difficulties encountered. It then proposes and adapts methods for simulating this type of system directly from its bond graph within a software package. The value of this study is in giving practical guidance to modellers, so that they can implement the adapted method in software.
Verschiedene intelligente Heimautomatisierungsgeräte wie Lampen, Schlösser und Thermostate verbreiten sich rasant im privaten Umfeld. Ein typisches Kommunikationsprotokoll für diese Geräteklasse ist Bluetooth Low Energy (BLE). In dieser Arbeit wird eine strukturierte Sicherheitsanalyse für BLE vorgestellt. Die beschriebene Vorgehensweise kategorisiert bekannte Angriffsvektoren und beschreibt einen möglichen Aufbau für eine Analyse. Im Zuge dieser Arbeit wurden einige sicherheitsrelevante Probleme aufgedeckt, die es Angreifern ermöglichen die Geräte vollständig zu übernehmen. Es zeigte sich, dass im Standard vorgesehene Sicherheitsfunktionen wie Verschlüsselung und Integritätsprüfungen häufig gar nicht oder fehlerhaft implementiert sind.
Surrogate models are used to reduce the burden of expensive-to-evaluate objective functions in optimization. By creating models which map genomes to objective values, these models can estimate the performance of unknown inputs, and so be used in place of expensive objective functions. Evolutionary techniques such as genetic programming or neuroevolution commonly alter the structure of the genome itself. A lack of consistency in the genotype is a fatal blow to data-driven modeling techniques: interpolation between points is impossible without a common input space. However, while the dimensionality of genotypes may differ across individuals, in many domains, such as controllers or classifiers, the dimensionality of the input and output remains constant. In this work we leverage this insight to embed differing neural networks into the same input space. To judge the difference between the behavior of two neural networks, we give them both the same input sequence, and examine the difference in output. This difference, the phenotypic distance, can then be used to situate these networks into a common input space, allowing us to produce surrogate models which can predict the performance of neural networks regardless of topology. In a robotic navigation task, we show that models trained using this phenotypic embedding perform as well or better as those trained on the weight values of a fixed topology neural network. We establish such phenotypic surrogate models as a promising and flexible approach which enables surrogate modeling even for representations that undergo structural changes.
The initially large number of variants is reduced by applying custom variant annotation and filtering procedures. This requires complex software toolchains to be set up and data sources to be integrated. Furthermore, increasing study sizes subsequently require higher efforts to manage datasets in a multi-user and multi-institution environment. It is common practice to expect numerous iterations of continuative respecification and refinement of filter strategies, when the cause for a disease or phenotype is unknown. Data analysis support during this phase is fundamental, because handling the large volume of data is not possible or inadequate for users with limited computer literacy. Constant feedback and communication is necessary when filter parameters are adjusted or the study grows with additional samples. Consequently, variant filtering and interpretation becomes time-consuming and hinders a dynamic and explorative data analysis by experts.
Are quality diversity algorithms better at generating stepping stones than objective-based search?
(2019)
The route to the solution of complex design problems often lies through intermediate "stepping stones" which bear little resemblance to the final solution. By greedily following the path of greatest fitness improvement, objective-based search overlooks and discards stepping stones which might be critical to solving the problem. Here, we hypothesize that Quality Diversity (QD) algorithms are a better way to generate stepping stones than objective-based search: by maintaining a large set of solutions which are of high-quality, but phenotypically different, these algorithms collect promising stepping stones while protecting them in their own "ecological niche". To demonstrate the capabilities of QD we revisit the challenge of recreating images produced by user-driven evolution, a classic challenge which spurred work in novelty search and illustrated the limits of objective-based search. We show that QD far outperforms objective-based search in matching user-evolved images. Further, our results suggest some intriguing possibilities for leveraging the diversity of solutions created by QD.
The initial phase in real world engineering optimization and design is a process of discovery in which not all requirements can be made in advance, or are hard to formalize. Quality diversity algorithms, which produce a variety of high performing solutions, provide a unique chance to support engineers and designers in the search for what is possible and high performing. In this work we begin to answer the question how a user can interact with quality diversity and turn it into an interactive innovation aid. By modeling a user's selection it can be determined whether the optimization is drifting away from the user's preferences. The optimization is then constrained by adding a penalty to the objective function. We present an interactive quality diversity algorithm that can take into account the user's selection. The approach is evaluated in a new multimodal optimization benchmark that allows various optimization tasks to be performed. The user selection drift of the approach is compared to a state of the art alternative on both a planning and a neuroevolution control task, thereby showing its limits and possibilities.
More and more devices will be connected to the internet [3]. Many devicesare part of the so-called Internet of Things (IoT) which contains many low-powerdevices often powered by a battery. These devices mainly communicate with the manufacturers back-end and deliver personal data and secrets like passwords.