Refine
H-BRS Bibliography
- yes (233) (remove)
Departments, institutes and facilities
- Fachbereich Wirtschaftswissenschaften (55)
- Fachbereich Informatik (53)
- Fachbereich Angewandte Naturwissenschaften (52)
- Institut für Technik, Ressourcenschonung und Energieeffizienz (TREE) (36)
- Fachbereich Ingenieurwissenschaften und Kommunikation (35)
- Fachbereich Sozialpolitik und Soziale Sicherung (33)
- Institut für Verbraucherinformatik (IVI) (18)
- Institute of Visual Computing (IVC) (14)
- Institut für funktionale Gen-Analytik (IFGA) (10)
- Institut für Sicherheitsforschung (ISF) (9)
Document Type
- Article (79)
- Conference Object (67)
- Part of a Book (32)
- Book (monograph, edited volume) (20)
- Report (10)
- Doctoral Thesis (8)
- Preprint (5)
- Master's Thesis (4)
- Contribution to a Periodical (2)
- Bachelor Thesis (1)
Year of publication
- 2019 (233) (remove)
Has Fulltext
- no (233) (remove)
Keywords
- Lehrbuch (4)
- Aminoacylase (2)
- Chemie (2)
- Design (2)
- Exergame (2)
- Extrusion blow molding (2)
- FPGA (2)
- Hyperspectral image (2)
- ICT (2)
- Innovation (2)
Traffic sign recognition is an important component of many advanced driving assistance systems, and it is required for full autonomous driving. Computational performance is usually the bottleneck in using large scale neural networks for this purpose. SqueezeNet is a good candidate for efficient image classification of traffic signs, but in our experiments it does not reach high accuracy, and we believe this is due to lack of data, requiring data augmentation. Generative adversarial networks can learn the high dimensional distribution of empirical data, allowing the generation of new data points. In this paper we apply pix2pix GANs architecture to generate new traffic sign images and evaluate the use of these images in data augmentation. We were motivated to use pix2pix to translate symbolic sign images to real ones due to the mode collapse in Conditional GANs. Through our experiments we found that data augmentation using GAN can increase classification accuracy for circular traffic signs from 92.1% to 94.0%, and for triangular traffic signs from 93.8% to 95.3%, producing an overall improvement of 2%. However some traditional augmentation techniques can outperform GAN data augmentation, for example contrast variation in circular traffic signs (95.5%) and displacement on triangular traffic signs (96.7 %). Our negative results shows that while GANs can be naively used for data augmentation, they are not always the best choice, depending on the problem and variability in the data.
Qualität als Erfolgsfaktor
(2019)
Analytische Chemie I
(2019)
Die unterschiedlichen Facetten der digitalen Zukunft zu beleuchten – sei es die aktive Gestaltungsaufgabe der Politik, die ethischen und moralischen Anpassungen durch Digitalisierung in der Gesellschaft oder die technische und wirtschaftliche Verantwortung – und in Bezug zueinander zu setzen, ist Aufgabe und Ziel dieser Publikation. Im Zuge der digitalen Transformation ist zudem der Ruf nach einer neuen Kultur für Gesellschaft, Politik und Wirtschaft geboten.
Meine Zeitung geht online
(2019)
The media is considered to be the fourth pillar in a democratic country. It acts as an effective control mechanism to check the other branches of the government. But this is only consequential when the media functions in an independent and transparent fashion with trained and neutral professionals who are aware of the accountability and consequences of their work. All these factors together would further the country as a democratic institution. Traditionally, it was legacy media responsible for a one-to-many communication process. Their goal was to provide information to the citizens. But this changed with development in technology and the use of social media in daily life. The internet brought with it new media formats which are easily accessible but also unstructured. These lowered barriers of entry in the media enabled citizens to become active participants in the communication process. As a result, these citizens developed a different relationship with the already existing media wherein they were not only the receivers to information but also co-producers. Real-time information allows users to communicate with each other and in turn widely generate public opinion on internet platforms. A many-to-many communication style emerged. While on the one hand, this type of discourse could be an opportunity for citizens to exercise their fundamental freedom of speech and expression, it is on the other hand, proving to have a detrimental effect in two parts: Lack of neutrality, polarized views and pre-existing misconceptions on the part of citizens as well as algorithms and formation of echo-chambers on the part of technology. Some questions arise in this scenario about the capability of citizen journalists, the duties they should adhere to along with the enjoyment of their rights and freedoms, the risks involved in an unchecked method of communication and the effect of citizen journalism in the democratic process.