Adversarial attacks on multi-task visual perception for autonomous driving. - In: The journal of imaging science & technology, ISSN 1943-3522, Bd. 65 (2021), 6, S. 60408-1-60408-9
In recent years, deep neural networks (DNNs) have accomplished impressive success in various applications, including autonomous driving perception tasks. However, current deep neural networks are easily deceived by adversarial attacks. This vulnerability raises significant - concerns, particularly in safety-critical applications. As a result, research into attacking and defending DNNs has gained much coverage. In this work, detailed adversarial attacks are applied on a diverse multi-task visual perception deep network across distance estimation, semantic segmentation, - motion detection, and object detection. The experiments consider both white and black box attacks for targeted and un-targeted cases, while attacking a task and inspecting the effect on all others, in addition to inspecting the effect of applying a simple defense method. We conclude this paper - by comparing and discussing the experimental results, proposing insights and future work. The visualizations of the attacks are available at https://youtu.be/6AixN90budY.
https://doi.org/10.2352/J.ImagingSci.Technol.2021.65.6.060408
Model-driven aspect-specific systems engineering in the automotive domain. - In: IEEE RASSE 2021, (2021), insges. 8 S.
The design and development of modern automobiles have become a big challenge for the automotive industry. The complexity of automotive hard-and software systems constantly increases due to the development and advancement of various kinds of safety, security, and comfort features. Even though various tools for model-based systems engineering exist in the automotive domain, some cover every phase and every aspect of the whole development process. The extent and complexity of the resulting models make it difficult to efficiently analyze, evaluate and possibly optimize corresponding architectures concerning specific aspects.An approach to overcome these barriers is the development of aspect-specific toolchains based on automotive architectural models. Such toolchains must be specially tailored to certain aspects of interest, and at the same time, be sufficiently adaptable to offer flexibility and reusability. Modern model-driven approaches can be used to achieve these goals.This paper presents a model-driven development workflow for aspect-specific tools for analyzing, evaluating, and optimizing specific measures of automotive hard-and software architectures. It presents some details of an aspect-specific application developed as a proof of concept for the suggested workflow. Moreover, it presents some challenges using the suggested workflow and the developed tool for a complex real-world automotive system model.
https://doi.org/10.1109/RASSE53195.2021.9686946
Model-driven application development for evaluation and optimization of automotive E/E-architectures. - In: IEEE RASSE 2021, (2021), insges. 8 S.
Over the last decades, automobiles have developed from predominantly mechanical machines to driving computers, consisting of a large number of sensors, actuators, and electronic control units that use various types of communication busses to form large and complex cyber-physical systems which provide a variety of comfort and safety features for the driver. Such E/E-systems (electric/electronic systems) do not only require high availability and reliability, but also overall efficient architectures and topologies. Because of the already high (and constantly increasing) level of complexity of such systems, the evaluation and optimization of their corresponding architectures has become a big challenge. The need for evaluation and optimization methods that are capable of abstracting the systems' complexity is evident for the automotive industry. In general, when it comes to designing and developing hard- and software systems, the paradigm of model-driven engineering emphasizes and supports the measures of abstraction, flexibility, and reusability to be able to grasp the complexity of modern systems. This paper presents a model-driven application for evaluating and optimizing automotive E/E-architectures as part of an integrated, model-based toolchain, developed with the Eclipse Modeling Framework.
https://doi.org/10.1109/RASSE53195.2021.9686943
Adaptive gain super-twisting-algorithm: design and discretization. - In: 60th IEEE Conference on Decision and Control, (2021), S. 6415-6420
In this paper, an eigenvalue-based discretization scheme is applied to a novel adaptive super-twisting-algorithm. Following the proposed procedure the discretization chattering effect is avoided entirely. An attractive property of the adaptation law is the insensitivity of the closed-loop system to overly large gains which in existing laws potentially leads to instability. Using Lyapunov's direct method the stability of the feedback loop is shown. Numerical examples underline the beneficial properties of the proposed methodology.
https://doi.org/10.1109/CDC45484.2021.9683304
Parameter identification of fractional-order LTI systems using modulating functions with memory reduction. - In: 60th IEEE Conference on Decision and Control, (2021), S. 5156-5162
The parameter estimation problem of a linear time-invariant fractional-order system is investigated by means of the modulating function method. Based on the assumption of known model structure and derivative orders, the modulating function method can be generalized to the fractional-order case in three different ways. We show that two approaches are identical for linear systems. This facilitates the computation of the fractional-order derivatives of modulating functions. In comparison to integer-order systems we have to include the initialization of the fractional-order system. We show that the spline type modulating function is capable of reducing the effect of the memory on the parameter estimation. However, it is not possible to compensate the memory initialization completely. In contrast to these tuning principles also the robustness against measurement noise must be considered. For this purpose we decouple the memory and noise compensation. The adjusted spline-type modulating functions reduce the initialization effect and the recursive least squares estimation provides the possibility to increase the numbers of equations such that the effect of the noise is reduced.
https://doi.org/10.1109/CDC45484.2021.9682920
Discrete-time implementation of an adaptive gain first-order sliding mode control law. - In: 60th IEEE Conference on Decision and Control, (2021), S. 6397-6402
A discrete-time implementation of a continuous-time adaptive gain sliding mode control law for a system with matched disturbance is presented. The discrete-time control algorithm is derived from the solution of the nominal continuous-time closed-loop dynamics. This approach ensures elimination of discretization chattering as well as proper disturbance rejection. Slight modifications of the resulting discrete-time control law are proposed to guarantee ultimate boundedness of the sliding variable and the adaptive gain which is formally proven by means of Lyapunov arguments. Prevention of discretization chattering and disturbance attenuation properties are validated in a simulation and compared to other approaches.
https://doi.org/10.1109/CDC45484.2021.9682922
Analog QRS detector with low power wireless communication module for medical IoT. - In: Biomedical engineering, ISSN 1862-278X, Bd. 66 (2021), S. S180
Enthalten in: 05-1700-P9
https://doi.org/10.1515/bmt-2021-6025
Realisierung eines Kopfphantoms für bio-elektrische Anwendungen. - In: Biomedical engineering, ISSN 1862-278X, Bd. 66 (2021), S. S348
Enthalten in: 07-0900-C1
https://doi.org/10.1515/bmt-2021-6054
Dry EEG recordings for detection of left, right hand, tongue and feet movements. - In: Biomedical engineering, ISSN 1862-278X, Bd. 66 (2021), S. S345
Enthalten in: 07-0900-C1
https://doi.org/10.1515/bmt-2021-6054
Real-time source space EEG time-frequency analysis. - In: Biomedical engineering, ISSN 1862-278X, Bd. 66 (2021), S. S34
Enthalten in: 05-0915-A6
https://doi.org/10.1515/bmt-2021-6006