1 - 10 of 275
-
PublicationConfocal fluorescence microscopy with high-NA diffractive lens arrays( 2022)Traditionally, there is a trade-off between the numerical aperture and field of view for a microscope objective. Diffractive lens arrays (DLAs) with overlapping apertures are used to overcome such a problem. A spot array with an NA up to 0.83 and a pitch of 75 m is produced by the proposed DLA at a wavelength of 488 nm. By measurement of the fluorescence beads, the DLA-based confocal setup shows the capability of high-resolution measurement over an area of 3mm 3mm with a 2.5 0.07 NA objective. Further, the proposed fluorescence microscope is insensitive to optical aberrations, which has been demonstrated by imaging with a simple doublet lens.
-
PublicationTowards a Better Understanding of Machine Learning based Network Intrusion Detection Systems in Industrial Networks( 2022)It is crucial in an industrial network to understand how and why a intrusion detection system detects, classifies, and reports intrusions. With the ongoing introduction of machine learning into the research area of intrusion detection, this understanding gets even more important since the used systems often appear as a black-box for the user and are no longer understandable in an intuitive and comprehensible way. We propose a novel approach to understand the internal characteristics of a machine learning based network intrusion detection system. This approach includes methods to understand which data sources the system uses, to evaluate whether the system uses linear or non-linear classification approaches, and to find out which underlying machine learning model is implemented in the system. Our evaluation on two publicly available industrial datasets shows that the detection of the data source and the differentiation between linear and non-linear models is possible with our approach. In addition, the identification of the underlying machine learning model can be accomplished with statistical significance for non-linear models. The information made accessible by our approach helps to develop a deeper understanding of the functioning of a network intrusion detection system, and contributes towards developing transparent machine learning based intrusion detection approaches.
-
PublicationCluster Crash: Learning from Recent Vulnerabilities in Communication Stacks( 2022)To ensure functionality and security of network stacks in Industrial Devices, thorough testing is necessary. This includes blackbox network fuzzing, where fields in network packets are filled with unexpected values to test the device's behavior in edge cases. Due to resource constraints, the tests need to be efficient and such the input values need to be chosen intelligently. Previous solutions use heuristics based on vague knowledge from previous projects to make these decisions. We aim to structure existing knowledge by defining Vulnerability Anti-Patterns for network communication stacks based on an analysis of the recent vulnerability groups Ripple20, Amnesia:33, and Urgent/11. For our evaluation, we implement fuzzing test scripts based on the Vulnerability Anti-Patterns and run them against 8 Industrial Devices from 5 different device classes. We show (I) that similar vulnerabilities occur in implementations of the same protocol as well as in different protocols, (II) that similar vulnerabilities also spread over different device classes, and (III) that test scripts based on the Vulnerability Anti-Patterns help to identify these vulnerabilities.
-
PublicationIntelligente Bild- und Videoauswertung für die Sicherheit( 2022)
-
PublicationMotorFactory: A Blender Add-on for Large Dataset Generation of Small Electric Motors( 2022)To enable automatic disassembly of different product types with uncertain condition and degree of wear in remanufacturing, agile production systems that can adapt dynamically to changing requirements are needed. Machine learning algorithms can be employed due to their generalization capabilities of learning from various types and variants of products. However, in reality, datasets with a diversity of samples that can be used to train models are difficult to obtain in the initial period. This may cause bad performances when the system tries to adapt to new unseen input data in the future. In order to generate large datasets for different learning purposes, in our project, we present a Blender add-on named MotorFactory to generate customized mesh models of various motor instances. MotorFactory allows to create mesh models which, complemented with additional add-ons, can be further used to create synthetic RGB images, depth images, normal images, segmentation ground truth masks and 3D point cloud datasets with point-wise semantic labels. The created synthetic datasets may be used for various tasks including motor type classification, object detection for decentralized material transfer tasks, part segmentation for disassembly and handling tasks, or even reinforcement learning-based robotics control or view-planning.
-
PublicationSovereign Digital Consent through Privacy Impact Quantification and Dynamic Consent( 2022)Digitization is becoming more and more important in the medical sector. Through electronic health records and the growing amount of digital data of patients available, big data research finds an increasing amount of use cases. The rising amount of data and the imposing privacy risks can be overwhelming for patients, so they can have the feeling of being out of control of their data. Several previous studies on digital consent have tried to solve this problem and empower the patient. However, there are no complete solution for the arising questions yet. This paper presents the concept of Sovereign Digital Consent by the combination of a consent privacy impact quantification and a technology for proactive sovereign consent. The privacy impact quantification supports the patient to comprehend the potential risk when sharing the data and considers the personal preferences regarding acceptance for a research project. The proactive dynamic consent implementation provides an implementation for fine granular digital consent, using medical data categorization terminology. This gives patients the ability to control their consent decisions dynamically and is research friendly through the automatic enforcement of the patients' consent decision. Both technologies are evaluated and implemented in a prototypical application. With the combination of those technologies, a promising step towards patient empowerment through Sovereign Digital Consent can be made.
-
PublicationFast and Lightweight Online Person Search for Large-Scale Surveillance Systems( 2022)The demand for methods for video analysis in the fieldof surveillance technology is rapidly growing due to theincreasing amount of surveillance footage available. Intelligent methods for surveillance software offer numerouspossibilities to support police investigations and crime prevention. This includes the integration of video processingpipelines for tasks such as detection of graffiti, suspiciousluggage, or intruders. Another important surveillance taskis the semi-automated search for specific persons-of-interestwithin a camera network. In this work, we identify the major obstacles for the development of person search systemsas the real-time processing capability on affordable hardware and the performance gap of person detection and reidentification methods on unseen target domain data. Inaddition, we demonstrate the current potential of intelligentonline person search by developing a real-world, largescale surveillance system. An extensive evaluation is provided for person detection, tracking, and re-identificationcomponents on affordable hardware setups, for which thewhole system achieves real-time processing up to 76 FPS.
-
PublicationWhere are we with Human Pose Estimation in Real-World Surveillance?( 2022)The rapidly increasing number of surveillance cameras offers a variety of opportunities for intelligent video analytics to improve public safety. Among many others, the automatic recognition of suspicious and violent behavior poses a key task. To preserve personal privacy, prevent ethnic bias, and reduce complexity, most approaches first extract the pose or skeleton of persons and subsequently perform activity recognition. However, current literature mainly focuses on research datasets and does not consider real-world challenges and requirements of human pose estimation. We close this gap by analyzing these challenges, such as inadequate data and the need for real-time processing, and proposing a framework for human pose estimation in uncontrolled crowded surveillance scenarios. Our system integrates mitigation measures as well as a tracking component to incorporate temporal information. Finally, we provide a detailed quantitative and qualitative analysis on both a scientific and a real-world dataset to highlight improvements and remaining obstacles towards robust real-world human pose estimation in uncooperative scenarios.
-
PublicationModelling Ambiguous Assignments for Multi-Person Tracking in Crowds( 2022)Multi-person tracking is often solved with a tracking-by-detection approach that matches all tracks and detections simultaneously based on a distance matrix. In crowded scenes, ambiguous situations with similar track-detection distances occur, which leads to wrong assignments. To mitigate this problem, we propose a new association method that separately treats such difficult situations by modelling ambiguous assignments based on the differences in the distance matrix. Depending on the numbers of tracks and detections, for which the assignment task is determined ambiguous, different strategies to resolve these ambiguous situations are proposed. To further enhance the performance of our tracking framework, we introduce a camera motion-aware interpolation technique and make an adaptation to the motion model, which improves identity preservation. The effectiveness of our approach is demonstrated through extensive ablative experiments with different detection models. Moreover, the superiority w.r.t. other trackers is shown on the challenging MOT17 and MOT20 datasets, where state-of-the-art results are obtained.
-
PublicationTowards Lower Precision Quantization for Pedestrian Detection in Crowded Scenario( 2021)Automatic pedestrian detection in real-world uncooperative scenarios is a well-known problem in computer vision, which has again gained in visibility last year due to distancing requirements. This remains a very challenging task, especially in crowded areas. Due to diverse technical and privacy issues, embedded systems such as smart cameras and smaller drones are becoming ubiquitous. Those complex detection models are not designed for on-edge processing in resource-constrained environments. Therefore, quantization techniques are required, in order to reduce the weights of a model to low-precision and not only effectively compress the model, but also allow to use low bit width arithmetic, which in term can be accelerated from specialized hardware. However, using an effective quantization scheme while maintaining accuracy is challenging. In this work we first establish a Quantization-aware training (QAT) and Post-training Quantization (PTQ) baseline for 8-bit uniform quantization to RetinaNet for person detection on the extremely challenging PANDA dataset. Those achieve near lossless performance in terms of accuracy by about 5× speedup of the CPU inference and 4× model size reduction for 8-bit PTQ quantized model. Further experiments with aggressive quantization scheme in 4- and 2-bit show diverse challenges resulting in severe instabilities. We apply both uniform and non-uniform quantization to overcome those and provide insights and strategies to fully quantize in 4- and 2-bit. Through this process we systematically evaluate the sensibility of individual parts of RetinaNet for quantization in very low precision. Finally, we show the resistance of quantization for limited amount of data.
