Refine
Year of publication
- 2024 (14)
- 2023 (28)
- 2022 (19)
- 2021 (26)
- 2020 (29)
- 2019 (44)
- 2018 (43)
- 2017 (37)
- 2016 (32)
- 2015 (30)
- 2014 (41)
- 2013 (43)
- 2012 (38)
- 2011 (40)
- 2010 (35)
- 2009 (38)
- 2008 (41)
- 2007 (38)
- 2006 (27)
- 2005 (25)
- 2004 (35)
- 2003 (27)
- 2002 (29)
- 2001 (31)
- 2000 (31)
- 1999 (31)
- 1998 (26)
- 1997 (33)
- 1996 (24)
- 1995 (22)
- 1994 (21)
- 1993 (18)
- 1992 (13)
- 1991 (14)
- 1990 (9)
- 1989 (17)
- 1988 (11)
- 1987 (8)
- 1986 (9)
- 1985 (9)
- 1984 (3)
- 1983 (7)
- 1982 (4)
- 1981 (3)
- 1980 (12)
- 1979 (6)
- 1978 (7)
- 1977 (1)
- 1976 (7)
- 1975 (3)
- 1974 (4)
- 1973 (1)
- 1972 (3)
- 1971 (2)
- 1970 (1)
Institute
- Fachbereich Elektrotechnik und Informationstechnik (1150) (remove)
Has Fulltext
- no (1150) (remove)
Document Type
- Article (629)
- Conference Proceeding (286)
- Book (112)
- Part of a Book (64)
- Patent (17)
- Report (9)
- Other (8)
- Conference: Meeting Abstract (6)
- Contribution to a Periodical (6)
- Doctoral Thesis (6)
Keywords
- Enterprise Architecture (5)
- MINLP (5)
- Engineering optimization (4)
- Digitale Transformation (3)
- Digitalisierung (3)
- Gamification (3)
- Literaturanalyse (3)
- Optimization (3)
- Powertrain (3)
- Referenzmodellierung (3)
With the many achievements of Machine Learning in the past years, it is likely that the sub-area of Deep Learning will continue to deliver major technological breakthroughs [1]. In order to achieve best results, it is important to know the various different Deep Learning frameworks and their respective properties. This paper provides a comparative overview of some of the most popular frameworks. First, the comparison methods and criteria are introduced and described with a focus on computer vision applications: Features and Uses are examined by evaluating papers and articles, Adoption and Popularity is determined by analyzing a data science study. Then, the frameworks TensorFlow, Keras, PyTorch and Caffe are compared based on the previously described criteria to highlight properties and differences. Advantages and disadvantages are compared, enabling researchers and developers to choose a framework according to their specific needs.
Through a mirror darkly – On the obscurity of teaching goals in game-based learning in IT security
(2021)
Teachers and instructors use very specific language communicating teaching goals. The most widely used frameworks of common reference are the Bloom’s Taxonomy and the Revised Bloom’s Taxonomy. The latter provides distinction of 209 different teaching goals which are connected to methods. In Competence Developing Games (CDGs - serious games to convey knowledge) and in IT security education, a two- or three level typology exists, reducing possible learning outcomes to awareness, training, and education. This study explores whether this much simpler framework succeeds in achieving the same range of learning outcomes. Method wise a keyword analysis was conducted. The results were threefold: 1. The words used to describe teaching goals in CDGs on IT security education do not reflect the whole range of learning outcomes. 2. The word choice is nevertheless different from common language, indicating an intentional use of language. 3. IT security CDGs use different sets of terms to describe learning outcomes, depending on whether they are awareness, training, or education games. The interpretation of the findings is that the reduction to just three types of CDGs reduces the capacity to communicate and think about learning outcomes and consequently reduces the outcomes that are intentionally achieved.
The increasing complexity of Advanced Driver Assistance Systems (ADAS) presents a challenging task to validate safe and reliable performance of these systems under varied conditions. The test and validation of ADAS/AD with real test drives, although important, involves huge costs and time. Simulation tools provide an alternative with the added advantage of reproducibility but often use ideal sensors, which do not reflect real sensor output accurately. This paper presents a new validation methodology using fault injection, as recommended by the ISO 26262 standard, to test software and system robustness. In our work, we investigated and developed a tool capable of inserting faults at different software and system levels to verify its robustness. The scope of this paper is to cover the fault injection test for the Visteon’s DriveCore™ system, a centralized domain controller for Autonomous driving which is sensor agnostic and SoC agnostic. With this new approach, the validation of safety monitoring functionality and its behavior can be tested using real-world data instead of synthetic data from simulation tools resulting in having better confidence in system performance before proceeding with in-vehicle testing.