Kennington C.,Universitatsstrasse 25 |
Schlangen D.,Universitatsstrasse 25
Computer Speech and Language | Year: 2017
Referring to visually perceivable objects is a very common occurrence in everyday language use. In order to produce expressions that refer, the speaker needs to be able to pick out visual properties that the referred object has and determine the words that name those properties, such that the expression can direct a listener's attention to the intended object. The speaker can aid the listener by looking in the direction of the object and by providing a pointing gesture to indicate it. In order to resolve the reference, the listener has a difficult job to do: simultaneously use all of the linguistic and non-linguistic information; the words of the referring expression that denote properties of the object, such as its colour or shape, need to already be known, and the non-linguistic gaze direction and pointing gesture of the speaker need to be incorporated. Crucially, the listener does not wait until the end of the referring expression before she begins to resolve it; rather, she is interpreting it as it unfolds. A model that resolves referring expressions as the listener must be able to do all of these things. In this paper, we present such a generative model of reference resolution. We explain our model and show empirically through a series of experiments that the model can work incrementally (i.e., word for word) as referring expressions unfold, can incorporate multimodal information such as gaze and pointing gestures in two ways, can learn a grounded meaning of words in the referring expression, can incorporate contextual (i.e., saliency) information, and is robust to noisy input such as automatic speech recognition transcriptions, as well as uncertainty in the representation of the candidate objects. © Published by Elsevier Ltd.
Aswolinskiy W.,Universitatsstrasse 25 |
Reinhart R.F.,Fraunhofer Research Institution for Mechatronic Systems Design IEM |
Steil J.J.,TU Braunschweig
Neurocomputing | Year: 2016
We consider the modelling of parametrized processes, where the goal is to model the process for new parameter value combinations. We compare the classical regression approach to a modular approach based on regression in the model space: First, for each process parametrization a model is learned. Second, a mapping from process parameters to model parameters is learned. We evaluate both approaches on two synthetic and two real-world data sets and show the advantages of the regression in the model space. © 2017 Elsevier B.V.
Burney S.,University Paris - Sud |
Wenzel R.,Free University of Berlin |
Kottke T.,Universitatsstrasse 25 |
Roussel T.,University Paris - Sud |
And 5 more authors.
Angewandte Chemie - International Edition | Year: 2012
A quick switch: A single amino acid substitution at a conserved residue (D396N) of Arabidopsis cryptochrome-1 (Atcry1) confers single-stranded DNA repair activity in vitro, conferring photolyase activity onto the cryptochrome (see graph). The mutant protein undergoes photoreduction of flavin to the fully reduced anionic form, similar to photolyases and unlike wild-type cryptochromes. © 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim.
Budke C.,Universitatsstrasse 25 |
Budke C.,Bielefeld University |
Dreyer A.,Universitatsstrasse 25 |
Dreyer A.,TU Hamburg - Harburg |
And 11 more authors.
Crystal Growth and Design | Year: 2014
Experimental investigations of ice recrystallization inhibition (IRI) efficacy have been performed for a large number of different substances, including natural antifreeze proteins (AFP) and antifreeze glycoproteins (AFGP), several synthetic AFGP analogues, as well as synthetic polymers. Here we define IRI efficacy as that concentration at which the ice recrystallization rate is dominated by the IRI compound. The investigated 39 compounds show IRI efficacies from about 2 mmol L-1 for the least effective compound still showing activity to about 1 nmol L-1, which corresponds to the highest efficacy found for natural AFGP samples. Hence, the assay employed allows for a quantitative comparison of IRI efficacy over a range of at least 6 orders of magnitude, thereby enabling studies of distinguishing effects induced by even subtle structural variations in AFGP analogues that were synthesized. Our results show that AFGP are by far the most effective IRI agents in our assay, and we surmise that this particular efficacy may be due to their disaccharide moieties. This supposition is supported by the fact that IRI efficacy is strongly reduced for monosaccharide AFGP analogues, as well as for AFGP analogues with acetyl-protected monosaccharide moieties. © 2014 American Chemical Society.
Shah A.B.,Bielefeld University |
Schielzeth H.,Bielefeld University |
Schielzeth H.,Friedrich - Schiller University of Jena |
Albersmeier A.,Universitatsstrasse 25 |
And 2 more authors.
Ecology and Evolution | Year: 2016
Despite recent advances in high-throughput sequencing, difficulties are often encountered when developing microsatellites for species with large and complex genomes. This probably reflects the close association in many species of microsatellites with cryptic repetitive elements. We therefore developed a novel approach for isolating polymorphic microsatellites from the club-legged grasshopper (Gomphocerus sibiricus), an emerging quantitative genetic and behavioral model system. Whole genome shotgun Illumina MiSeq sequencing was used to generate over three million 300 bp paired-end reads, of which 67.75% were grouped into 40,548 clusters within RepeatExplorer. Annotations of the top 468 clusters, which represent 60.5% of the reads, revealed homology to satellite DNA and a variety of transposable elements. Evaluating 96 primer pairs in eight wild-caught individuals, we found that primers mined from singleton reads were six times more likely to amplify a single polymorphic microsatellite locus than primers mined from clusters. Our study provides experimental evidence in support of the notion that microsatellites associated with repetitive elements are less likely to successfully amplify. It also reveals how advances in high-throughput sequencing and graph-based repetitive DNA analysis can be leveraged to isolate polymorphic microsatellites from complex genomes. © 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd.
Kuck D.,Bielefeld University |
Kuck D.,Universitatsstrasse 25 |
Linke J.,Bielefeld University |
Teichmann L.C.,Bielefeld University |
And 6 more authors.
Physical Chemistry Chemical Physics | Year: 2016
The solid-state molecular structure of centrohexaindane (1), a unique hydrocarbon comprising six benzene rings clamped to each other in three dimensions around a neopentane core, and the molecular packing in crystals of 1·CHCl3 are reported. The molecular Td-symmetry and the Cartesian orientation of the six indane wings of 1 in the solid state have been confirmed. The course and limitation of electrophilic aromatic substitution of 1 are demonstrated for the case of nitration. Based on nitration experiments of a lower congener of 1, tribenzotriquinacene 5, the six-fold nitrofunctionalisation of 1 has been achieved in excellent yield, giving four constitutional isomers, two nonsymmetrical (14 and 17) and two C3-symmetrical ones (15 and 16), all of which contain one single nitro group in each of the six benzene rings. The relative yields of the four isomers (∼3:1:1:3) point to a random electrophilic attack of the electrophiles at the twelve formally equivalent outer positions of the aromatic periphery of 1, suggesting electronic independence of its six aromatic π-electron systems. In turn, the pronounced conformational rigidity of the centrohexacyclic framework of 1 enables the unequivocal structural identification of the isomeric hexanitrocentrohexaindanes 14-17 by 1H NMR spectroscopy. © 2016 the Owner Societies.
Neubauer K.,University of Rostock |
Huhns M.,University of Rostock |
Hausmann T.,University of Rostock |
Klemke F.,Humboldt University of Berlin |
And 6 more authors.
Journal of Biotechnology | Year: 2012
A chimeric cyanophycin synthetase gene composed of the cphA Te coding region from the cyanobacterium Thermosynechococcus elongatus BP-1, the constitutive 35S promoter and the plastid targeting sequence of the integral photosystem II protein PsbY was transferred to the tobacco variety Petit Havanna SRI and the commercial potato starch production variety Albatros. The resulting constitutive expression of cyanophycin synthetase leads to polymer contents in potato leaf chloroplasts of up to 35mg/g dry weight and in tuber amyloplasts of up to 9mg/g dry weight. Both transgenic tobacco and potato were used for the development of isolation methods applicable for large-scale extraction of the polymer. Two different procedures were developed which yielded polymer samples of 80 and 90% purity, respectively. © 2011 Elsevier B.V.
Lier F.,CITEC |
Siepmann F.,CITEC |
Paul-Stueve T.,CITEC |
Wrede S.,Universitatsstrasse 25 |
And 2 more authors.
ACM International Conference Proceeding Series | Year: 2012
Researchers and other knowledge workers frequently produce and use diverse research artefacts such as papers, data sets, experiment specifications, software, etc. In this, they are often faced with unclear relationships (e.g., which version of a software was in use for a particular paper), creating unnecessary work and potentially errors. Semantic web technologies can provide metadata as well as explicit, specific links between the artefacts. However, data acquisition and perceived utility are potential stumbling blocks for adoption. Therefore, we propose a system which is focused on integrating and augmenting existing data (thus protecting the existing investment), and examine it using an interactionoriented perspective, on users without semantic web experience. Specifically, we first study requirements of the target group and then present an exploratory study of managing research artefacts related to software-centric projects. The results confirm that diverse data sources are in common use, that re-using existing repositories is perceived as efficient (e.g., more convenient, shorter cycle time), and that the experimented aggregates are perceived as functionally relevant. Furthermore, the integration of quality assurance mechanisms, such as continuous integration, is perceived as beneficial, despite some added effort. Copyright 2012 ACM.
Lier F.,CITEC |
Schulz S.,CITEC |
Lutkebohle I.,Universitatsstrasse 25
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2012
Simulated environments often provide the first, and are usually the most frequent, test environment for robotic systems, primarily due to their cost and safety advantages. Unfortunately, changing aspects of both, the simulation and the real robot, as well as actuator control algorithms are often not taken into account when relying on simulation results. In this paper we present a continuous integration approach to verify simulated robot models in an integrated and frequent manner, comprising a simulated and a real robot for comparison. The central aspect of our concept is to iteratively assess the fidelity of simulated robot models. In an exemplary case study we distilled a first set of requirements and metrics, which can be used by developers to verify their algorithms and to automatically detect further system changes. © 2012 Springer-Verlag Berlin Heidelberg.
Gorzolka K.,Universitatsstrasse 25 |
Bednarz H.,Universitatsstrasse 25 |
Niehaus K.,Universitatsstrasse 25
Planta | Year: 2014
Mass spectrometry imaging was applied on germinated barley for the detailed localization of metabolites in longitudinal and transversal seed sections. Among others, 20 m/z signals occurred in three regular peak clusters with specific, distinct localizations in embryo tissues. High resolution FT-ICR MS, MALDI-TOF MS/MS, and UHPLC-ESI MS/MS served for the identification and structural characterization of these compounds. Only five metabolites were published in their structures, namely the antifungal compounds hordatine A and B in non-glycosylated and glycosylated forms. All other non-identified cluster compounds were of hordatine-like structure and differed by systematic O-methylations, hydroxylations, and glycosylations. These differences in molecular structures correlated to distinct localization patterns within the embryo and might serve for the regulation of antifungal properties. Based on the structural investigations by mass spectrometry, an array of different hordatines that comprises the five published hordatines, 15 novel hordatine derivates and their six precursors could be localized in the embryo of germinated barley. Implications for the biosynthetic pathway and transport processes are discussed. © 2014 Springer-Verlag Berlin Heidelberg.