Klaus Janschek
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Klaus Janschek has a total of 50 co-authors. Here are the top co-authors by number of shared publications:
- Kristof Richmond12 shared publications
- Andrey Morozov
Institute of Automation Technische Universität Dresden Dresden Germany
6 shared publications - Kai Ding
Technische Universität Dresden
3 shared publications Albert H. Sultanov
Ufa State Aviation Technical Univ. (Russia)
2 shared publications- Chao Yao
Autonomous Robotics Research Center (ARRC), Technology Innovation Institute (TII), Abu Dhabi, United Arab Emirates
2 shared publications - Christel Baier
Technische Universität Dresden
2 shared publications
Latest Publications
- Open Access
- 1 March 2021
- Published by IOP Publishing in IOP Conference Series: Materials Science and Engineering
- Vol. 1069 (1) , 012010
Evaluation of graphic data corruptions impact on artificial intelligence applications
The article is devoted to the problem of the reliability of applications based on artificial intelligence. The authors made an attempt to evaluate the impact of the graphic data distortion at the input of a convolutional neural network on the result of image classification. The experiment is based on the fault injection method. A series of independent tests were carried out for such distortions as Gaussian noise, salt and pepper, speckle and Poisson noise, as well as median blur, motion blur, scene brightness changing, rotation, rain, and snow. The results showed that Gaussian noise was the least critical distortion; environmental conditions (rain, snow, brightness) and image rotation up to 20 degrees are less critical than focus losing and motion blur, while the most critical distortion is speckle noise. It was verified that preprocessing the input data of the neural network improves the accuracy of image recognition.
- 1 March 2021
- Published by IOP Publishing in IOP Conference Series: Materials Science and Engineering
- Vol. 1069 (1) , 012010
Open AccessEvaluation of graphic data corruptions impact on artificial intelligence applications
The article is devoted to the problem of the reliability of applications based on artificial intelligence. The authors made an attempt to evaluate the impact of the graphic data distortion at the input of a convolutional neural network on the result of image classification. The experiment is based on the fault injection method. A series of independent tests were carried out for such distortions as Gaussian noise, salt and pepper, speckle and Poisson noise, as well as median blur, motion blur, scene brightness changing, rotation, rain, and snow. The results showed that Gaussian noise was the least critical distortion; environmental conditions (rain, snow, brightness) and image rotation up to 20 degrees are less critical than focus losing and motion blur, while the most critical distortion is speckle noise. It was verified that preprocessing the input data of the neural network improves the accuracy of image recognition.
- Open Access
- 26 April 2020
- Published by Open Publishing Association in Electronic Proceedings in Theoretical Computer Science
- Vol. 316, 121-133
Iterative Variable Reordering: Taming Huge System Families
For the verification of systems using model-checking techniques, symbolic representations based on binary decision diagrams (BDDs) often help to tackle the well-known state-space explosion problem. Symbolic BDD-based representations have been also shown to be successful for the analysis of families of systems that arise, e.g., through configurable parameters or following the feature-oriented modeling approach. The state space of such system families face an additional exponential blowup in the number of parameters or features. It is well known that the order of variables in ordered BDDs is crucial for the size of the model representation. Especially for automatically generated models from real-world systems, family models might even be not constructible due to bad variable orders. In this paper we describe a technique, called iterative variable reordering, that can enable the construction of large-scale family models. We exemplify feasibility of our approach by means of an aircraft velocity control system with redundancy mechanisms modeled in the input language of the probabilistic model checker PRISM. We show that standard reordering and dynamic reordering techniques fail to construct the family model due to memory and time constraints, respectively, while the new iterative approach succeeds to generate a symbolic family model.
- 26 April 2020
- Published by Open Publishing Association in Electronic Proceedings in Theoretical Computer Science
- Vol. 316, 121-133
Open AccessIterative Variable Reordering: Taming Huge System Families
For the verification of systems using model-checking techniques, symbolic representations based on binary decision diagrams (BDDs) often help to tackle the well-known state-space explosion problem. Symbolic BDD-based representations have been also shown to be successful for the analysis of families of systems that arise, e.g., through configurable parameters or following the feature-oriented modeling approach. The state space of such system families face an additional exponential blowup in the number of parameters or features. It is well known that the order of variables in ordered BDDs is crucial for the size of the model representation. Especially for automatically generated models from real-world systems, family models might even be not constructible due to bad variable orders. In this paper we describe a technique, called iterative variable reordering, that can enable the construction of large-scale family models. We exemplify feasibility of our approach by means of an aircraft velocity control system with redundancy mechanisms modeled in the input language of the probabilistic model checker PRISM. We show that standard reordering and dynamic reordering techniques fail to construct the family model due to memory and time constraints, respectively, while the new iterative approach succeeds to generate a symbolic family model.
- 1 January 2020
- Published by Elsevier in IFAC-PapersOnLine
- Vol. 53 (2) , 9420-9425
Stereo Visual-Inertial Fusion for UAV State Estimation
- 1 January 2020
- Published by Elsevier in IFAC-PapersOnLine
- Vol. 53 (2) , 9420-9425
Stereo Visual-Inertial Fusion for UAV State Estimation
- 1 January 2020
- Published by Elsevier in IFAC-PapersOnLine
- Vol. 53 (2) , 8915-8922
Model-Based Dependability Assessment of Phased-Mission Unmanned Aerial Vehicles
- 1 January 2020
- Published by Elsevier in IFAC-PapersOnLine
- Vol. 53 (2) , 8915-8922
Model-Based Dependability Assessment of Phased-Mission Unmanned Aerial Vehicles
- 1 January 2019
- Published by Research Publishing Services
Breaking the Limits of Redundancy Systems Analysis
- 1 January 2019
- Published by Research Publishing Services
Breaking the Limits of Redundancy Systems Analysis
- 12 December 2018
- Published by Elsevier in IFAC-PapersOnLine
- Vol. 51 (22) , 115-120
Modeling and Sliding Mode Control of a Fully-actuated Multirotor with Tilted Propellers
- 12 December 2018
- Published by Elsevier in IFAC-PapersOnLine
- Vol. 51 (22) , 115-120
Modeling and Sliding Mode Control of a Fully-actuated Multirotor with Tilted Propellers
- 11 October 2018
- Published by Elsevier in IFAC-PapersOnLine
- Vol. 51 (24) , 1417-1424
Stochastic Model-based Analysis of Timing Errors for Mechatronic Systems with User-defined General Discrete-time Distributions
- 11 October 2018
- Published by Elsevier in IFAC-PapersOnLine
- Vol. 51 (24) , 1417-1424
Stochastic Model-based Analysis of Timing Errors for Mechatronic Systems with User-defined General Discrete-time Distributions
- 11 October 2018
- Published by Elsevier in IFAC-PapersOnLine
- Vol. 51 (24) , 1067-1074
SysML-based Profile for Dependable UAV Design
- 11 October 2018
- Published by Elsevier in IFAC-PapersOnLine
- Vol. 51 (24) , 1067-1074
SysML-based Profile for Dependable UAV Design
- 1 August 2018
- Published by American Institute of Aeronautics and Astronautics (AIAA) in Journal of Guidance, Control, and Dynamics
- Vol. 41 (8) , 1776-1782
Monocamera-Based Solar Radiation Pressure Estimation for Small-Body Rendezvous Missions
- 1 August 2018
- Published by American Institute of Aeronautics and Astronautics (AIAA) in Journal of Guidance, Control, and Dynamics
- Vol. 41 (8) , 1776-1782
Monocamera-Based Solar Radiation Pressure Estimation for Small-Body Rendezvous Missions
- 1 December 2017
- Published by Walter de Gruyter GmbH in at - Automatisierungstechnik
- Vol. 65 (12) , 838-850
Entwicklung von Modelltransformationen für multimodale Benutzungsschnittstellen
Zusammenfassung: Mit dem sogenannten Model-based User Interface Development existiert ein vielversprechendes Konzept zur automatischen Erzeugung von Benutzungsschnittstellen. Die Entwicklung der hierfür benötigten Modelltransformationen stellt allerdings einen aufwändigen, fehleranfälligen und daher kostenintensiven Prozess dar. Zur Erleichterung dieses Prozesses schlägt dieser Beitrag eine spezielle Vorgehensweise vor, die auf der Konfiguration von Modelltransformationen in deklarativen Abbildungsmodellen basiert und die explizite Berücksichtigung von Kontextabhängigkeiten ermöglicht.
- 1 December 2017
- Published by Walter de Gruyter GmbH in at - Automatisierungstechnik
- Vol. 65 (12) , 838-850
Entwicklung von Modelltransformationen für multimodale Benutzungsschnittstellen
Zusammenfassung: Mit dem sogenannten Model-based User Interface Development existiert ein vielversprechendes Konzept zur automatischen Erzeugung von Benutzungsschnittstellen. Die Entwicklung der hierfür benötigten Modelltransformationen stellt allerdings einen aufwändigen, fehleranfälligen und daher kostenintensiven Prozess dar. Zur Erleichterung dieses Prozesses schlägt dieser Beitrag eine spezielle Vorgehensweise vor, die auf der Konfiguration von Modelltransformationen in deklarativen Abbildungsmodellen basiert und die explizite Berücksichtigung von Kontextabhängigkeiten ermöglicht.