Search

Manipulation of granular materials such as sand or rice remains an unsolved problem due to challenges such as the difficulty of defining their configuration or modeling the materials and their particles interactions. Current approaches tend to simplify the material dynamics and omit the interactions between the particles. In this paper, we propose to use a graph-based representation to model the interaction dynamics of the material and rigid bodies manipulating it. This allows the planning of manipulation trajectories to reach a desired configuration of the material. We use a graph neural network (GNN) to model the particle interactions via message-passing. To plan manipulation trajectories, we propose to minimise the Wasserstein distance between a predicted distribution of granular particles and their desired configuration. We demonstrate that the proposed method is able to pour granular materials into the desired configuration both in simulated and real scenarios., Comment: 8 pages, 5 figures. Accepted to IEEE Robotics and Automation Letters (RA-L)

This article introduces AI2D-RST, a multimodal corpus of 1000 English-language diagrams that represent topics in primary school natural sciences, such as food webs, life cycles, moon phases and human physiology. The corpus is based on the Allen Institute for Artificial Intelligence Diagrams (AI2D) dataset, a collection of diagrams with crowd-sourced descriptions, which was originally developed to support research on automatic diagram understanding and visual question answering. Building on the segmentation of diagram layouts in AI2D, the AI2D-RST corpus presents a new multi-layer annotation schema that provides a rich description of their multimodal structure. Annotated by trained experts, the layers describe (1) the grouping of diagram elements into perceptual units, (2) the connections set up by diagrammatic elements such as arrows and lines, and (3) the discourse relations between diagram elements, which are described using Rhetorical Structure Theory (RST). Each annotation layer in AI2D-RST is represented using a graph. The corpus is freely available for research and teaching., Comment: 24 pages; revised version submitted to Language Resources & Evaluation

Efficient index structures for fast approximate nearest neighbor queries are required in many applications such as recommendation systems. In high-dimensional spaces, many conventional methods suffer from excessive usage of memory and slow response times. We propose a method where multiple random projection trees are combined by a novel voting scheme. The key idea is to exploit the redundancy in a large number of candidate sets obtained by independently generated random projections in order to reduce the number of expensive exact distance evaluations. The method is straightforward to implement using sparse projections which leads to a reduced memory footprint and fast index construction. Furthermore, it enables grouping of the required computations into big matrix multiplications, which leads to additional savings due to cache effects and low-level parallelization. We demonstrate by extensive experiments on a wide variety of data sets that the method is faster than existing partitioning tree or hashing based approaches, making it the fastest available technique on high accuracy levels.

This paper describes the academic library environment in Finland, discussing increases in research funding that have not been directly evident in library budgets. It then describes the FinELib program, the National Electronic Library, which has a centralized approach to joint acquisition of electronic information resources. The key issues have been agreed upon by all members of the university consortia after which negotiations and other operations can be carried out in a flexible manner. The funding principles of the FinELib Consortium 1998-2002 are outlined for resource type, multidisciplinary material, and subject specific material. The remainder of the paper discusses the future of the National Electronic Library, network and information security, and current challenges of the FinELib Consortium model. (AEF)

This paper describes the design and implementation of a networked professional development initiative for information literacy, run by DEDICATE (Distance EDucation Information Courses with Access Through nEtworks), a European Union-funded project under the Telematics for Libraries 4th Framework Program. Library and academic staff from five Central and Eastern European universities participated in the course, which is based on principles of collaborative and experiential online learning and has as its goal the development by participants of information literacy courses tailored to the needs of user groups within each of the five partner universities. The paper describes the pedagogic approach; main features of the DEDICATE course and its technological environment; and implementation and evaluation. (Contains 15 references.) (Author/AEF)

Significant concentrations of carbidopa (CD) were found in rat hypothalamus, striatum, and in striatal microdialysis efflux after intraperitoneal administration of the drug. Efflux levels peaked one hour after administration of 100 mg/kg at 0.37 micrograms/ml, or about 2% of serum levels. Concurrent CD levels in hypothalamus and striatum were about 2.5% and 1.5%, respectively, of corresponding serum levels. Levels of dopamine and its principal metabolites in striatal efflux were unaffected. The removal of the brain blood by saline perfusion decreased the striatal and hypothalamic CD concentrations only by 33% and 16%, respectively. In other rats receiving both CD and levodopa (LD), brain L-dopa, dopamine and 3,4-dihydroxyphenylacetic acid (DOPAC) levels after one hour tended to be proportionate to LD dose. When the LD dose remained constant, increasing the CD dose dose-dependently enhanced L-dopa levels in the hypothalamus and striatum. However dopamine levels did not increase but, in contrast, decreased dose-dependently (although significantly only in the hypothalamus). CD also caused dose-dependent decrease in striatal 3-O-methyldopa (3-OMD) and in striatal and hypothalamic homovanillic acid (HVA), when the LD dose was 50 mg/kg. We conclude that, at doses exceeding 50 mg/kg, sufficient quantities of CD enter the brain to inhibit dopamine formation, especially in the hypothalamus. Moreover, high doses of LD/CD, both of which are themselves catechols, can inhibit the O-methylation of brain catecholamines formed from the LD.

Significant concentrations of carbidopa (CD) were found in rat hypothalamus, striatum, and in striatal microdialysis efflux after intraperitoneal administration of the drug. Efflux levels peaked one hour after administration of 100 mg/kg at 0.37 microg/kg or about 2 percent of serum levels. Concurrent CD levels in hypothalamus and striatum were about 2.5 percent and 1.5 percent, respectively, of corresponding serum levels. Levels of dopamine and its principal metabolites in striatal efflux were unaffected. The removal of the brain blood by saline perfusion decreased the striatal and hypothalamic CD concentrations only by 33 percent and 16 percent, respectively. In other rats receiving both CD and levodopa (LD), brain L-dopa, dopamine, and 3,4-dihydroxyphenvlacetic acid (DOPAC) levels after one hour tended to be proportionate to LD dose. When the LD dose remained constant, increasing the CD dose dose-dependently enhanced L-dopa levels in the hypothalamus and striatum. However, dopamine levels did not increase but, in contrast, decreased dose-dependently (although significantly only in the hypothalamus). CD also caused dose-dependent decrease in striatal 3-O-methyldopa (3-OMD) and in striatal and hypothalamic homovanillic acid (HVA), when the LD dose was 50 mg/kg. We conclude that, at doses exceedimg 50 mg/kg, sufficient quantities of CD enter the brain to inhibit dopamine formation, especially in the hypothalamus. Moreover, high doses of LD/CD, both of which are themselves catechols, can inhibit the O-methylation of brain catecholamines formed from the LD.