1 - 10 of 63
-
-
-
PublicationSecurity in Industrial Environments( 2021)Industrial Control Systems (ICS) are used to control and monitor industrial environments such as production or chemical processes. Due to their elementary role in critical infrastructure, ICS pose an interesting target for attacker. This is why it is necessary to secure ICS, to close vulnerabilities, and to make them more robust against attacks. In this talk, we will dive into the art of industrial ethical hacking, vulnerability scanning, and fuzzing. This will come along with an overview of the state-of-the-art in the field of industrial security. In addition, this talk will give information on different recent attacks and vulnerabilities such as TRITON and Ripple20.
-
PublicationSET-260: A Measurement Campaign for EO/IR Signatures of UAVs( 2021)
-
PublicationSmart City Technologies for Cultural Heritage ProtectionClimate change (CC) will morph the environmental landscape, thus leading to climate stress imposed on Cultural Heritage (CH). Especially, tangible CH, like castles, palaces, monuments and churches as well as gardens are exposed to CH effects. Such effects are heat waves, flooding, higher sea level, just to name a few. The management and preservation of such CH buildings and whole sites, particularly in the context of CC, is a complex task in which authorities and decision makers need to aggregate and oversee information from diverse sources and domains. Yet, only by considering all relevant and available information, stakeholders can make well-grounded decisions. This imposes a complex task upon the authorities, not only due to the diversity and heterogeneity, but also to the quantity of available data. Only if the current and future situation of the CH in focus is understood, strategies for protecting them can be developed. The first challenge is to apply different kind of sensors to the buildings and gardens to collect data about the weather (temperature, precipitation, etc.), the situation of walls incl. cracks and the state of plants. After that, this data needs to be managed and made accessible in homogeneous way for further processing and analysis. The domain of smart city research faces the exact same problems. Sensors are applied all over the city for example about traffic, infrastructure, air and water quality and weather data. In contrast to CH the community is much larger and the industry is involved as well. Therefore, it is beneficial to look into technologies developed for smart cities and analyze how they can be applied to the monitoring of CH sites. For retrieving, managing and processing sensor data there are open standards evolving, for example the SensorThings API standard by the Open Geospatial Consortium (OGC). Currently, many tools evolve around such standards from which some are available as open source. First results of successfully applying these technologies from different CH and smart city projects will be presented.
-
PublicationEmerging Perspectives on Medical Decision Support: Co-Designing XAI( 2020)Artificial Intelligence (AI) and Machine Learning (ML) promise to significantly enhance the capabilities of decision support systems in medicine. Yet, if these systems fail to providean understandable rationale of the decision making process the adoption of this powerful technology will be difficult. Hence, there is growing interest in Explainable Artificial Intelligence (XAI). Explanations that speak the language of the user are a cornerstone for Fair AI. In this position paper, we reflect on research about XAI and designing decision support in medicine. From there, we lay out an approach to use co-design methodology to explore how users perceive and interact with explanations of decision support systems.
-
PublicationA Short Note on Analyzing Sequence Complexity in Trajectory Prediction Benchmarks( 2020)The analysis and quantification of sequence complexity is an open problem frequently encountered when defining trajectory prediction benchmarks. In order to enable a more informative assembly of a data basis, an approach for determining a dataset representation in terms of a small set of distinguishable prototypical sub-sequences is proposed. The approach employs a sequence alignment followed by a learning vector quantization (LVQ) stage. A first proof of concept on synthetically generated and real-world datasets shows the viability of the approach.
-
-
PublicationInvestigation on Combining 3D Convolution of Image Data and Optical Flow to Generate Temporal Action Proposals( 2019)In this paper, several variants of two-stream architectures for temporal action proposal generation in long, untrimmed videos are presented. Inspired by the recent advances in the field of human action recognition utilizing 3D convolutions in combination with two-stream networks and based on the Single-Stream Temporal Action Proposals (SST) architecture [3], four different two-stream architectures utilizing sequences of images on one stream and sequences of images of optical flow on the other stream are subsequently investigated. The four architectures fuse the two separate streams at different depths in the model; for each of them, a broad range of parameters is investigated systematically as well as an optimal parametrization is empirically determined. The experiments on the THUMOS' 14 [11] dataset - containing untrimmed videos of 20 different sporting activities for temporal action proposals - show that all four two-stream architectures are able to outperform the original single-stream SST and achieve state of the art results. Additional experiments revealed that the improvements are not restricted to one method of calculating optical flow by exchanging the method of Brox [1] with FlowNet2 [10] and still achieving improvements.
-
PublicationGeneral Fail-Safe Emergency Stopping for Highly Automated Vehicles( 2019)From SAE level 3 onwards, automated vehicles must be able to resolve sudden system failures without driver intervention, including failure modes that are difficult or impossible to address by redundancy alone. Causes of hazardous multiple-point faults-beyond internal failures-include lightning strikes or deliberate attacks by electromagnetic pulses. Stopping the vehicle under such conditions is challenging: A full braking maneuver may risk rear-end collisions or loss of traction; likewise, any other constant braking profile will pose considerable risk of not achieving a true ""safe state"". This paper presents an emergency stopping system to execute a situation-dependent braking maneuver that can resolve system failures up to(but not limited to) a full electrics/electronics failure, with the aim of providing a baseline safety solution for all failure modes (short of mechanical failures) for which no dedicated solution is available. The system is composed of an electronic planning unit and a hydraulic/mechanical subsystem, both of which are implemented and tested in simulated and in real environments.
