Fraunhofer-Institut für Optronik, Systemtechnik und Bildauswertung IOSB

Filters
Reset filters

Settings
Now showing 1 - 10 of 24
  • Publication
    On Residual-based Diagnosis of Physical Systems
    ( 2022)
    Diedrich, Alexander
    ;
    Niggemann, Oliver
    In this article we describe a novel diagnosis methodology for physical systems such as industrial production systems. The article consists of two parts: Part one analyzes the differences between using sensor values and using residual values for fault diagnosis. Residual values denote the health of a component by comparing sensor values to a predefined model of normal behaviour. We further analyse how faults propagate through components of a physical system and argue for the use of residual values for diagnosing physical systems. In part two we extend the theory of established consistency-based diagnosis algorithms to use residual values. We also illustrate how users of the presented diagnosis methodology are free to substitute the residual generating equations and the diagnosis algorithm to suit their specific needs. For diagnosis, we present the algorithm HySD, based on Satisfiability Modulo Linear Arithmetic. We present an implementation of HySD using threshold values and a symbolic diagnosis approach. However, the approach is also suitable to integrate modern machine learning methods for anomaly detection and combine them with a multitude of diagnosis approaches. Through experiments on the process-industry benchmark Tennessee Eastman Process and another benchmark consisting of multiple tank systems we show the feasibility of our approach. Overall we show how our novel diagnosis approach offers a practical methodology that allows industry to advance from current state of the art anomaly detection to automated fault diagnosis.
  • Publication
    Application of LSTM Networks for Water Demand Prediction in Optimal Pump Control
    ( 2021)
    Kühnert, Christian
    ;
    Gonuguntla, Naga
    ;
    Krieg, Helene
    ;
    Nowak, Dimitri
    ;
    Thomas, Jorge A.
    Every morning, water suppliers need to define their pump schedules for the next 24 h for drinking water production. Plans must be designed in such a way that drinking water is always available and the amount of unused drinking water pumped into the network is reduced. Therefore, operators must accurately estimate the next day's water consumption profile. In real-life applications with standard consumption profiles, some expert system or vector autoregressive models are used. Still, in recent years, significant improvements for time series prediction have been achieved through special deep learning algorithms called long short-term memory (LSTM) networks. This paper investigates the applicability of LSTM models for water demand prediction and optimal pump control and compares LSTMs against other methods currently used by water suppliers. It is shown that LSTMs outperform other methods since they can easily integrate additional information like the day of the week or national holidays. Furthermore, the online- and transfer-learning capabilities of the LSTMs are investigated. It is shown that LSTMs only need a couple of days of training data to achieve reasonable results. As the focus of the paper is on the real-world application of LSTMs, data from two different water distribution plants are used for benchmarking. Finally, it is shown that the LSTMs significantly outperform the system currently in operation.
  • Publication
    Are you sure? Prediction revision in automated decision-making
    ( 2021)
    Burkart, Nadia
    ;
    Robert, Sebastian
    ;
    Huber, Marco
    With the rapid improvements in machine learning and deep learning, decisions made by automated decision support systems (DSS) will increase. Besides the accuracy of predictions, their explainability becomes more important. The algorithms can construct complex mathematical prediction models. This causes insecurity to the predictions. The insecurity rises the need for equipping the algorithms with explanations. To examine how users trust automated DSS, an experiment was conducted. Our research aim is to examine how participants supported by an DSS revise their initial prediction by four varying approaches (treatments) in a between-subject design study. The four treatments differ in the degree of explainability to understand the predictions of the system. First we used an interpretable regression model, second a Random Forest (considered to be a black box [BB]), third the BB with a local explanation and last the BB with a global explanation. We noticed that all participants improved their predictions after receiving an advice whether it was a complete BB or an BB with an explanation. The major finding was that interpretable models were not incorporated more in the decision process than BB models or BB models with explanations.
  • Publication
    New active learning algorithms for near-infrared spectroscopy in agricultural applications
    ( 2021)
    Krause, Julius
    ;
    Günder, Maurice
    ;
    Schulz, Daniel
    ;
    Gruna, Robin
    The selection of training data determines the quality of a chemometric calibration model. In order to cover the entire parameter space of known influencing parameters, an experimental design is usually created. Nevertheless, even with a carefully prepared Design of Experiment (DoE), redundant reference analyses are often performed during the analysis of agricultural products. Because the number of possible reference analyses is usually very limited, the presented active learning approaches are intended to provide a tool for better selection of training samples.
  • Publication
    Multimedia analysis platform for crime prevention and investigation. Results of MAGNETO project
    ( 2021)
    Perez, Francisco J.
    ;
    Garrido, Victor J.
    ;
    Garcia, Alberto
    ;
    Zambrano, Marcelo
    ;
    Kozik, Rafal
    ;
    Choras, Michal
    ;
    Mühlenberg, Dirk
    ;
    Pallmer, Dirk
    ;
    Müller, Wilmuth
    Nowadays, the use of digital technologies is promoting three main characteristics of information, i.e. the volume, the modality and the frequency. Due to the amount of information generated by tools and individuals, it has been identified a critical need for the Law Enforcement Agencies to exploit this information and carry out criminal investigations in an effective way. To respond to the increasing challenges of managing huge amounts of heterogeneous data generated at high frequency, the paper outlines a modular approach adopted for the processing of information gathered from different information sources, and the extraction of knowledge to assist criminal investigation. The proposed platform provides novel technologies and efficient components for processing multimedia information in a scalable and distributed way, allowing Law Enforcement Agencies to make the analysis and a multidimensional visualization of criminal information in a single and secure point.
  • Publication
    Discretization of hybrid CPPS data into timed automaton using restricted Boltzmann machines
    ( 2020)
    Hranisavljevic, Nemanja
    ;
    Maier, Alexander
    ;
    Niggemann, Oliver
    Cyber-Physical Production Systems (CPPSs) are hybrid systems composed of a discrete and continuous part. However, most of the applied machine learning algorithms handle the dynamics of the two parts separately and in different fashions: for the discrete part, the notion of discrete events (and their timings) is essential (e.g. when learning automata or rules), while the dynamics of the continuous part is often defined by differential equations or time-series models. Reconciling the different nature of the two is a major challenge for machine learning. One solution is to express continuous behavior in discrete terms, i.e. the explicit events are extracted. Then, at the cost of information loss caused by discretization, the overall behavior can be jointly analyzed. This paper proposes a novel machine learning discretization approach called DENTA (Deep Network Timed Automaton) which solves the aforementioned challenges through the construction of an (overall) deterministic timed automaton from the original hybrid data. First, it hierarchically extracts new features from the continuous data using a deep network of stacked restricted Boltzmann machines (RBMs). We show that high-level RBM abstractions can further be used to automatically detect meaningful discrete events in continuous system behavior. Finally, a discrete representation of overall system behavior in the form of a timed automaton is created, which allows a joint timing analysis of the whole system. The model is verified by the anomaly detection on a synthetic and a real-world dataset and the results show clear advantages of the approach for a specific class of systems.
  • Publication
    Learning and Tracking the 3D Body Shape of Freely Moving Infants from RGB-D sequences
    ( 2020)
    Hesse, Nikolas
    ;
    Pujadis, Sergi
    ;
    Black, Michael J.
    ;
    Arens, Michael
    ;
    Hofmann, Ulrich
    ;
    Schroeder, Sebastian
    Statistical models of the human body surface are generally learned from thousands of high-quality 3D scans in predefined poses to cover the wide variety of human body shapes and articulations. Acquisition of such data requires expensive equipment, calibration procedures, and is limited to cooperative subjects who can understand and follow instructions, such as adults. We present a method for learning a statistical 3D Skinned Multi-Infant Linear body model (SMIL) from incomplete, low-quality RGB-D sequences of freely moving infants. Quantitative experiments show that SMIL faithfully represents the RGB-D data and properly factorizes the shape and pose of the infants. To demonstrate the applicability of SMIL, we fit the model to RGB-D sequences of freely moving infants and show, with a case study, that our method captures enough motion detail for General Movements Assessment (GMA), a method used in clinical practice for early detection of neurodevelopmental disorders in infants. SMIL provides a new tool for analyzing infant shape and movement and is a step towards an automated system for GMA.
  • Publication
    The Virtual Caliper: Rapid Creation of Metrically Accurate Avatars from 3D Measurements
    ( 2019)
    Pujades, Sergi
    ;
    Mohler, Betty
    ;
    Thaler, Anne
    ;
    Tesch, Joachim
    ;
    Mahmood, Naureen
    ;
    Hesse, Nikolas
    ;
    Bülthoff, Heinrich H.
    ;
    Black, Michael J.
    Creating metrically accurate avatars is important for many applications such as virtual clothing try-on, ergonomics, medicine, immersive social media, telepresence, and gaming. Creating avatars that precisely represent a particular individual is challenging however, due to the need for expensive 3D scanners, privacy issues with photographs or videos, and difficulty in making accurate tailoring measurements. We overcome these challenges by creating "The Virtual Caliper", which uses VR game controllers to make simple measurements. First, we establish what body measurements users can reliably make on their own body. We find several distance measurements to be good candidates and then verify that these are linearly related to 3D body shape as represented by the SMPL body model. The Virtual Caliper enables novice users to accurately measure themselves and create an avatar with their own body shape. We evaluate the metric accuracy relative to ground truth 3D body scan data, compare the method quantitatively to other avatar creation tools, and perform extensive perceptual studies. We also provide a software application to the community that enables novices to rapidly create avatars in fewer than five minutes. Not only is our approach more rapid than existing methods, it exports a metrically accurate 3D avatar model that is rigged and skinned.
  • Publication
    State estimation for tracking in image space with a de- and re-coupled IMM filter
    ( 2018)
    Becker, Stefan
    ;
    Hübner, Wolfgang
    ;
    Arens, Michael
    Estimating the motion state of objects is a central component of most visual tracking pipelines. Visual tracking relies on observations in scale space generated by an appearance model. Under real-life conditions, it is obvious to assume that the dynamic of a tracked object changes over time. A popular solution for considering such varying system characteristics is the Interacting Multiple Model (IMM) filter. Usually, the motion of objects is modeled using position, velocity, and acceleration. Although it seems obvious that different image space dimensions can be combined in one overall system state, this naïve approach may fail under various circumstances. Toward this end, we demonstrate the benefit of decoupling the state estimate of an IMM filter in case of relying solely on the output of a visual tracker. Further, a state re-coupling scheme is introduced which helps to better deal with the corresponding measurement uncertainties of such a tracking pipeline. The proposed decoupled and re-coupled IMM filters are evaluated on publicly available datasets.
  • Publication
    Fast algorithm for 2D fragment assembly based on partial EMD
    ( 2017)
    Zhang, M.
    ;
    Chen, S.
    ;
    Shu, Z.
    ;
    Xin, S.-Q.
    ;
    Zhao, J.
    ;
    Jin, G.
    ;
    Zhang, R.
    ;
    Beyerer, Jürgen
    2D Fragment assembly is an important research topic in computer vision and pattern recognition, and has a wide range of applications such as relic restoration and remote sensing image processing. The key to this problem lies in utilizing contour features or visual cues to find the optimal partial matching. Considering that previous algorithms are weak in predicting the best matching configuration of two neighboring fragments, we suggest using the earth moverâs distance, based on length/property correspondence, to measure the similarity, which potentially matches a point on the first contour to a desirable destination point on the second contour. We further propose a greedy algorithm for 2D fragment assembly by repeatedly assembling two neighboring fragments into a composite one. Experimental results on map-piece assembly and relic restoration show that our algorithm runs fast, is insensitive to noise, and provides a novel solution to the fragment assembly problem.