Johannes Kepler University of Linz

About: Johannes Kepler University of Linz is a education organization based out in Linz, Oberösterreich, Austria. It is known for research contribution in the topics: Thin film & Quantum dot. The organization has 6605 authors who have published 19243 publications receiving 385667 citations.

Relevance, discourse and the ‘hot’ stable core social representations—A structural analysis of word associations

Abstract: This paper presents an investigation of the structure of word associations dependent on the context in which they are assessed. Respondents from Spain and Nicaragua produced free associations about war and peace. Word associations about each of the two stimulus words were produced either spontaneously or within the context of a distracting priming condition in contrast to the association task. The semantic space for each stimulus word (war, peace) is analysed to find substructures of words which remain stable across contexts. These substructures or stable cores are taken to indicate a well-structured social representation as opposed to a loosely organized knowledge domain. Such cores were found for associations about war in both countries, but for peace in the Nicaraguan sample only. This finding is interpreted as a consequence of public discourse and symbolic coping with relevant or threatening objects or phenomena. Stable cores were found to consist primarily of ‘hot’ words, i.e. words which are proximal to an individual's experience. More intellectual and distant (‘cold’) words did not enter the stable core. Results are discussed in terms of the central core theory of social representations and of numerical consensus being an insufficient criterion for social representations.

Towards Accurate Generative Models of Video: A New Metric & Challenges

Abstract: Recent advances in deep generative models have lead to remarkable progress in synthesizing high quality images. Following their successful application in image processing and representation learning, an important next step is to consider videos. Learning generative models of video is a much harder task, requiring a model to capture the temporal dynamics of a scene, in addition to the visual presentation of objects. While recent attempts at formulating generative models of video have had some success, current progress is hampered by (1) the lack of qualitative metrics that consider visual quality, temporal coherence, and diversity of samples, and (2) the wide gap between purely synthetic video data sets and challenging real-world data sets in terms of complexity. To this extent we propose Frechet Video Distance (FVD), a new metric for generative models of video, and StarCraft 2 Videos (SCV), a benchmark of game play from custom starcraft 2 scenarios that challenge the current capabilities of generative models of video. We contribute a large-scale human study, which confirms that FVD correlates well with qualitative human judgment of generated videos, and provide initial benchmark results on SCV.

Central Moment Discrepancy (CMD) for Domain-Invariant Representation Learning

Abstract: The learning of domain-invariant representations in the context of domain adaptation with neural networks is considered. We propose a new regularization method that minimizes the discrepancy between domain-specific latent feature representations directly in the hidden activation space. Although some standard distribution matching approaches exist that can be interpreted as the matching of weighted sums of moments, e.g. Maximum Mean Discrepancy (MMD), an explicit order-wise matching of higher order moments has not been considered before. We propose to match the higher order central moments of probability distributions by means of order-wise moment differences. Our model does not require computationally expensive distance and kernel matrix computations. We utilize the equivalent representation of probability distributions by moment sequences to define a new distance function, called Central Moment Discrepancy (CMD). We prove that CMD is a metric on the set of probability distributions on a compact interval. We further prove that convergence of probability distributions on compact intervals w.r.t. the new metric implies convergence in distribution of the respective random variables. We test our approach on two different benchmark data sets for object recognition (Office) and sentiment analysis of product reviews (Amazon reviews). CMD achieves a new state-of-the-art performance on most domain adaptation tasks of Office and outperforms networks trained with MMD, Variational Fair Autoencoders and Domain Adversarial Neural Networks on Amazon reviews. In addition, a post-hoc parameter sensitivity analysis shows that the new approach is stable w.r.t. parameter changes in a certain interval. The source code of the experiments is publicly available.

Direct-bandgap emission from hexagonal Ge and SiGe alloys

Abstract: Silicon crystallized in the usual cubic (diamond) lattice structure has dominated the electronics industry for more than half a century. However, cubic silicon (Si), germanium (Ge) and SiGe alloys are all indirect-bandgap semiconductors that cannot emit light efficiently. The goal1 of achieving efficient light emission from group-IV materials in silicon technology has been elusive for decades2–6. Here we demonstrate efficient light emission from direct-bandgap hexagonal Ge and SiGe alloys. We measure a sub-nanosecond, temperature-insensitive radiative recombination lifetime and observe an emission yield similar to that of direct-bandgap group-III–V semiconductors. Moreover, we demonstrate that, by controlling the composition of the hexagonal SiGe alloy, the emission wavelength can be continuously tuned over a broad range, while preserving the direct bandgap. Our experimental findings are in excellent quantitative agreement with ab initio theory. Hexagonal SiGe embodies an ideal material system in which to combine electronic and optoelectronic functionalities on a single chip, opening the way towards integrated device concepts and information-processing technologies. A hexagonal (rather than cubic) alloy of silicon and germanium that has a direct (rather than indirect) bandgap emits light efficiently across a range of wavelengths, enabling electronic and optoelectronic functionalities to be combined on a single chip.

Abstract robust coarse spaces for systems of PDEs via generalized eigenproblems in the overlaps

Abstract: Coarse spaces are instrumental in obtaining scalability for domain decomposition methods for partial differential equations (PDEs). However, it is known that most popular choices of coarse spaces perform rather weakly in the presence of heterogeneities in the PDE coefficients, especially for systems of PDEs. Here, we introduce in a variational setting a new coarse space that is robust even when there are such heterogeneities. We achieve this by solving local generalized eigenvalue problems in the overlaps of subdomains that isolate the terms responsible for slow convergence. We prove a general theoretical result that rigorously establishes the robustness of the new coarse space and give some numerical examples on two and three dimensional heterogeneous PDEs and systems of PDEs that confirm this property.