Fakultät für Informatik
Refine
Year of publication
- 2010 (8) (remove)
Document Type
Language
- English (8)
Has Fulltext
- no (8)
Is part of the Bibliography
- no (8)
Keywords
Institute
The number of embedded systems in our daily lives that are distributed, hidden, and ubiquitous continues to increase. Many of them are safety-critical. To provide additional or better functionalities, they are becoming more and more complex, which makes it difficult to guarantee safety. It is undisputed that safety must be considered before the start of development, continue until decommissioning, and is particularly important during the design of the system and software architecture. An architecture must be able to avoid, detect, or mitigate all dangerous failures to a sufficient degree. For this purpose, the architectural design must be guided and verified by safety analyses. However, state-of-the-art component-oriented or model-based architectural design approaches use different levels of abstraction to handle complexity. So, safety analyses must also be applied on different levels of abstraction, and it must be checked and guaranteed that they are consistent with each other, which is not supported by standard safety analyses. In this paper, we present a consistency check for CFTs that automatically detects commonalities and inconsistencies between fault trees of different levels of abstraction. This facilitates the application of safety analyses in top-down architectural designs and reduces effort.
Efficient safety analyses of complex software intensive embedded systems are still a challenging task. This article illustrates how model-driven development principles can be used in safety engineering to reduce cost and effort. To this end, the article shows how well accepted safety engineering approaches can be shifted to the level of model-driven development by integrating safety models into functional development models. Namely, we illustrate how UML profiles, model transformations, and techniques for multi language development can be used to seamlessly integrate component fault trees into the UML.
Online Identification of Learner Problem Solving Strategies Using Pattern Recognition Methods
(2010)
Learning and programming environments used in computer science education give feedback to the users by system messages. These are triggered by programming errors and give only "technical" hints without regard to the learners' problem solving process. To adapt the messages not only to the factual but also to the procedural knowledge of the learners, their problem solving strategies have to be identified automatically and in process. This article describes a way to achieve this with the help of pattern recognition methods. Using data from a study with 65 learners aged 12 to 13 using a learning environment for programming, a classification system based on hidden Markov models is trained and integrated in the very same environment. We discuss findings in that data and the performance of the automatic online identification, and present first results using the developed software in class.
Reverberation effects as observed by room microphones severely degrade the performance of automatic speech recognition systems. We investigate the use of dereverberation by spectral subtraction as proposed by Lebart and Boucher and introduce a simple approach to estimate the required decay parameter by clapping hands. Experiments on small vocabulary continuous speech recognition task on read speech show that using the calibrated dereverberation improves WER from 73.2 to 54.7 for the best microphone. In combination with system adaptation, the WER could be reduced to 28.2, which is only a 16% relative loss of performance comparison to using a headset instead of a room microphone.
The CALO Meeting Assistant (MA) provides for distributed meeting capture, annotation, automatic transcription and semantic analysis of multiparty meetings, and is part of the larger CALO personal assistant system. This paper presents the CALO-MA architecture and its speech recognition and understanding components, which include real-time and offline speech transcription, dialog act segmentation and tagging, topic identification and segmentation, question-answer pair identification, action item recognition, decision extraction, and summarization.
Data fusion plays a central role in more and more automotive applications, especially for driver assistance systems. On the one hand the process of data fusion combines data and information to estimate or predict states of observed objects.
On the other hand data fusion introduces abstraction layers for data description and allows building more flexible and modular systems.The data fusion process can be divided into a low-level processing (tracking and object discrimination) and a high level processing (situation assessment).
High level processing becomes more and more the focus of current research as different assistance applications will be combined into one comprehensive assistance system.
Different levels/strategies for data fusion can be distinguished: Fusion on raw data level, fusion on feature level and fusion on decision level.
All fusion strategies can be found in current driver assistance implementations.
The paper gives an overview of the different fusion strategies and shows their application in current driver assistance systems. For low level processing a raw data fusion approach in a stereo video system is described, as an example for feature level fusion the fusion of radar and camera data for tracking is explained.
As an example for a high level fusion algorithm an approach for a situation assessment based on multiple sensors is given. The paper describes practical realizations of these examples and points out their potential to further increase traffic safety with reasonably low cost for the overall system.