Projects

 

Spiking - Spiking neural network
Research Program on the Computing and learning algorithms with spiking neural network.
Start: 19.10.2018
End: 18.10.2021
INRC - Principled development of drastically improved architectures and algorithms for computing and learning with spiking neurons on Loihi
We will apply biologically inspired ideas for drastically improving the computational performance and learning capability of recurrent networks of spiking neurons on Loihi through the introduction of working memory capabilities. In addition, we will introduce new principles for creating and improving computing and learning performance of these networks through Learning-to-Learn.
Start: 31.08.2018
End: 30.08.2021
FWF - VORONOI++ - Circle Expansion and abstract Voronoi diagrams
This project is concerned with a versatile and influential data structure called the Voronoi diagram, a geometric structure which makes explicit the proximity information exerted by a given set of sites in space. Space partitioning structures of this kind have proven useful not only in computational geometry and more applied areas of computer science, but also in the natural and economical sciences. Fast construction methods and, as a prerequisite, a thorough understanding of their structural and algorithmic properties, are in demand. In this DACH project, we intend to join forces to conduct research on some of these problems. The involved research groups (R. Klein, Bonn; E. Papadopoulou, Lugano; B. Jüttler, Linz; F. Aurenhammer, Graz) have successfully worked on this topic within the framework of EuroGIGA (initiated by F. Aurenhammer) in the Collaborative Research Project VORONOI'', which is documented by numerous relevant publications. Our main goal is to generalize Voronoi diagrams to such an extent that modeling real life scenarios becomes possible. The progress we have already made in previous collaborations has put this goal within our reach. Among our planned research topics are Abstract Voronoi diagrams, cluster Voronoi diagrams, anisotropic diagrams, and skeletal structures in 3D. These topics show the necessary diversity for a successful research and, on the other hand, are strongly interrelated which promises a (continuing) fruitful cooperation between the project partners. Complementing the planned theoretical research, practical aspects will be emphasized. The complexity of the structures to be investigated has reached a level where visualization tools (like interactive applets) are needed, which are intended to be made public later on. To put the findings of this project to practical use, software implementations of the developed algorithms for anisotropic Voronoi diagrams and 3D straight skeletons will be available.
Start: 31.05.2015
End: 30.05.2020
EU - HBP SGA2 - Human Brain Project Specific 2
Understanding the human brain is one of the greatest scientific challenges of our time. Such an understanding can provide profound insights into our humanity, leading to fundamentally new computing technologies, and transforming the diagnosis and treatment of brain disorders. Modern ICT brings this prospect within reach. The HBP Flagship Initiative (HBP) thus proposes a unique strategy that uses ICT to integrate neuroscience data from around the world, to develop a unified multi-level understanding of the brain and diseases, and ultimately to emulate its computational capabilities. The goal is to catalyze a global collaborative effort. During the HBP’s first Specific Grant Agreement (SGA1), the HBP Core Project will outline the basis for building and operating a tightly integrated Research Infrastructure, providing HBP researchers and the scientific Community with unique resources and capabilities. Partnering Projects will enable independent research groups to expand the capabilities of the HBP Platforms, in order to use them to address otherwise intractable problems in neuroscience, computing and medicine in the future. In addition, collaborations with other national, European and international initiatives will create synergies, maximizing returns on research investment. This document describes the HBP’s plans for SGA1, and details what steps will be taken to move the HBP closer to achieving its ambitious Flagship Objectives.
Start: 31.03.2018
End: 30.03.2020
FWF-SASNN - Stochastic Assemblies in Spiking Neural Networks
Recent experimental results have provided valuable insights in the organization of computations in biological neuronal networks. In particular, evidence for two main features of cortical computation is rapidly accumulating. First, neurons operate in concert with other neurons in so-called cell assemblies. Second, the activity of single neurons, synapses, and assemblies in the brain is highly stochastic. These findings force us to rethink how computations are organized in cortical neuronal networks. However, an integrated view on stochasticity and assembly organization in spiking neural networks is still missing. In this project, we will investigate stochastic assembly organization both in organic and artificial spiking neural networks. One emphasis will be the characterization of stochastic assembly activation in highly controllable setups and assembly emergence through plasticity processes. Experiments will be accompanied by theoretical modeling, analysis, and computer simulations that will help to understand the basic mechanisms that give rise to assembly formation. In particular, the investigations in this project will focus on (a) the characterization of stochastic assemblies their emergence through plasticity processes in cultured neural networks and acute brain slices, (b) the control of stochastic assemblies in cultured neural networks and its application to neuroprosthetics, and (c) computations in artificial spiking neural networks based on emergent stochastic assemblies with applications to novel computing and learning devices.
Start: 28.02.2017
End: 27.02.2020
EU-HBP SGA1 - Human Brain Project Specific Grant Agreement 1
Understanding the human brain is one of the greatest scientific challenges of our time. Such an understanding can provide profound insights into our humanity, leading to fundamentally new computing technologies, and transforming the diagnosis and treatment of brain disorders. Modern ICT brings this prospect within reach. The HBP Flagship Initiative (HBP) thus proposes a unique strategy that uses ICT to integrate neuroscience data from around the world, to develop a unified multi-level understanding of the brain and diseases, and ultimately to emulate its computational capabilities. The goal is to catalyze a global collaborative effort. During the HBP’s first Specific Grant Agreement (SGA1), the HBP Core Project will outline the basis for building and operating a tightly integrated Research Infrastructure, providing HBP researchers and the scientific Community with unique resources and capabilities. Partnering Projects will enable independent research groups to expand the capabilities of the HBP Platforms, in order to use them to address otherwise intractable problems in neuroscience, computing and medicine in the future. In addition, collaborations with other national, European and international initiatives will create synergies, maximizing returns on research investment. This document describes the HBP’s plans for SGA1, and details what steps will be taken to move the HBP closer to achieving its ambitious Flagship Objectives.
Start: 31.03.2016
End: 30.03.2018
EU-HBP - The Human Brain Project
The goal of the Human Brain Project, part of the FET Flagship Programme, is to translate this vision into reality, using ICT as a catalyst for a global collaborative effort to understand the human brain and its diseases and ultimately to emulate its computational capabilities.
Start: 30.09.2013
End: 30.03.2016
FWF - PNEUMA - Novel computational paradigms for memristive architectures
During the past two decades, philosophers, psychologists, cognitive scientists, clinicians and neuroscientists strived to provide authoritative definitions of consciousness within a neurobiological framework. Engineers have more recently joined this quest by developing neuromorphic VLSI circuits for emulating biological functions. Yet, to date artificial systems have not been able to faithfully recreate natural attributes such as true processing locality (memory and computation) and complexity (1010 synapses per cm2), preventing the achievement of a long-term goal: the creation of autonomous cognitive systems. This project aspires to develop experimental platforms capable of perceiving, learning and adapting to stimuli by leveraging on the latest developments of five leading European institutions in neuroscience, nanotechnology, modeling and circuit design. The non-linear dynamics as well as the plasticity of the newly discovered memristor are shown to support Spike-based- and Spike-Timing-Dependent-Plasticity (STDP), making this extremely compact device an excellent candidate for realizing large-scale self-adaptive circuits; a step towards autonomous cognitive systems. The intrinsic properties of real neurons and synapses as well as their organization in forming neural circuits will be exploited for optimizing CMOS-based neurons, memristive grids and the integration of the two into realtime biophysically realistic neuromorphic systems. Finally, the platforms would be tested with conventional as well as abstract methods to evaluate the technology and its autonomous capacity.
Start: 31.08.2011
End: 30.08.2015
EU - BRAINSCALES - Brain-inspired multiscale computation in neuromorphic hybrid systems
The BrainScaleS project aims at understanding function and interaction of multiple spatial and temporal scales in brain information processing. The fundamentally new approach of BrainScaelS lies in the in-vivo biological experimentation and computational analysis. Spatial scales range from individual neurons over larger neuron populations to entire functional brain areas. Temporal scales range from milliseconds relevant for event based plasticity mechanisms to hours or days relevant for learning and development. In the project generic theoretical principles will be extracted to enable an artificial synthesis of cortical-like cognitive skills. Both, numerical simulations on peta-flop supercomputers and fundamentally different non-von Neumann hardware architecture will be employed for this purpose. Neurobiological data from the early perceptual visual and somato-sensory systems will be combined with data from specifically targeted higher cortical areas. Functional databases as well as novel project-specific experimental tools and protocols will be developed and used. New theoretical concepts and methods will be developed for understanding the computational role of the complex multi-scale dynamics of neural systems in-vivo. Innovative in-vivo experiments will be carried out to guide this analytical understanding. Multiscale architectures will be synthesized into a non-von Neumann computing device realized in custom designed electronic hardware. The proposed Hybrid Multi-scale Computing Facility (HMF) combines microscopic neuro-morphic physical model circuits with numerically calculated meso-scopic and macroscopic functional units and a virtual environment providing sensory, decision-making and motor interfaces. The project also plans to employ peta-flop supercomputing to obtain new insights into the specific properties of the different hardware architectures. A set of demonstration experiments will link multi-scale analysis of biological systems with functionally and architecturally equivalent synthetic systems and offer the possibility for quantitative statements on the validity of theories bridging multiple scales. The demonstration experiments will also explore non-von Neumann computing outside the realm of brain-science. BrainScaelS will establish close links with the EU Brain-i-Nets and the Blue Brain project at the EPFL Lausanne. The consortium consists of a core group of 10 partners with 13 individual groups. Together with other projects and groups the BrainScaelS consortium plans to make important contributions to the preparation of a future FET flagship project. This project will address the understanding and exploitation of information processing in the human brain as one of the major intellectual challenges of humanity with vast potential applications.
Start: 31.12.2010
End: 30.12.2014
FWF - EuroGIGA - Advanced Voronoi (Delaunay) Structures
This CRP is devoted to the investigation of the geometric and algorithmic properties of geometric graphs, in particular, those arising from spatial decompositions. This is an interesting and demanding topic, with many aspects and applications, and remarkable relevant work has been done by the international scientific community, particularly in Europe. This CRP is meant to unite these strong forces, to provide a fruitful and sucessfull platform for a thorough investigation of this topic.The CRP is planned to consist of nine IPs from six European countries (Austria, Belgium, Germany, Poland, Spain, Switzerland), which include those where the tradition on the topics in this CRP is strong. Four main structures will be investigated in this CRP: Voronoi diagrams, Skeletal structures, Variants of triangulations, and Proximity graphs. These structures are interrelated, in a twofold sense. On the one hand, there is a direct geometric link between them, stemming from their definition as distance-based spatial structures. On the other, they share common features which allow their joint investigation on a higher level. Both aspects bear high potential of successful collaboration between the (groups of) people who by and large worked on these structures individually so far.
Start: 14.06.2011
End: 13.06.2014
EU - AMARSI - Adaptive Modular Architecture for Rich Motor Skills
Motor skills of humans and animals are still utterly astonishing when compared to robots. AMARSi aims at a qualitative jump in robotic motor skills towards biological richness.
AMARSi focuses on:
* compliant mechanics
* morphological computation
* human and robotic motor control: comparison
* damage-robust, safe, fast and flexible motion
* learning (unsupervised, reinforcement and imitation learning)
* dynamical neural models
* unifying locomotion and manipulation
AMARSi' objective: the experimentation and demonstration of rich motor skills on the iCub humanoid robot and on the quadruped Cheetah. Rich motor skills in robots will have a tremendous impact on our society. Dexterous and skillful motion in robots will make them more suitable for a large number of tasks. The compliant and natural movements will make them blend into everyday routines, safe and psychologically acceptable.
Start: 28.02.2010
End: 27.02.2014
EU - FACETS-ITN - Fast Analog Computing with Emergent Transient States - Initial Training Network
FACETS-ITN is a research and training network involving partners from 11 universities and research centres, 3 industrial companies and 1 semi-industrial research centre from 6 European countries. It combines competencies in neurobiology, computational neuroscience, information science, physics and electrical engineering. The scientific goal is to experimentally and theoretically explore the structure and the computational principles of biological neural circuits using in-vitro and in-vivo neurobiological experiments as well as analytical approaches, model building and simulation techniques. The concepts of learning and plasticity will be of particular importance. Based on the input from biology and modelling it is expected to prepare the grounds for novel hardware based computing devices that make use of such principles. Such devices will be built and operated in the form of large scale demonstrators as part of the research plan. Within the training network 21 selected Ph.D. students will be integrated into an existing international research environment and receive a strongly interdisciplinary training. The training comprises an intense exchange and visiting programme, specific training workshops for all scientific areas covered as well as in non-scientific key competencies. This training concept will enhance their original academic education in order to cope with the challenges of this diverse international research environment. The proposed training programme will be closely coupled to existing research projects as well as graduate programmes of participating universities. It will create a sustained infrastructure based on web-based learning as well as on scientific interdisciplinary networks and the intersectional exchange with industry. The community building among Ph.D. students from different disciplines will include students from other projects, in particular the current FACETS integrated project and its planned successor.
Start: 31.08.2009
End: 30.08.2013
EU - BRAIN-I-NETS - Novel Brain-Inspired Learning Paradigms for Large-Scale Neuronal Networks
Current designs of neurally inspired computing systems rely on learning rules that appear to be insufficient to port the superior adaptive and computational capabilities of biological neural systems into large-scale recurrent neural hardware system. This is not surprising, since most of these learning rules had to be extrapolated from results of neurobiological experiments in vitro. New experimental techniques in neurobiology such as 2-photon laser-scanning microscopy, optogenetic cell activation, and dynamic clamp techniques make it now possible to record the changes that really take place in the intact brain during learning. First results indicate that the rules for synaptic plasticity have in fact to be rewritten. In particular, it appears that local synaptic plasticity is gated in multiple ways by global factors such as neuromodulators and network states. One primary goal of this project is to apply and extend new cutting-edge experimental techniques to produce a set of rules for synaptic plasticity and network reorganisation that describe the actual adaptive processes that take place in the living brain during learning.

This new rules will be analysed by computational neuroscience experts and their consequences for learning in simulated large-scale networks of neurons and neurally inspired computing systems will be ascertained. The goal of this project is to port essential aspects of learning in the intact brain into current and next-generation neuromorphic hardware. New interchangeable software tools, that have recently been developed in the FP6 project FACETS, will be employed to carry out these investigations. Open questions that arise in these modelling studies will be addressed by changes in experimental protocols of the neuroscientists, building on long standing interdisciplinary collaborations among the partners.
Start: 31.12.2009
End: 30.12.2012
ORGANIC - Self-organized recurrent neural learning for language processing
The human brain is an unrivalled "engine" for speech processing and language understanding. It integrates a large variety of learning, adaptation, optimization and self-stabilization mechanisms across many dynamically interacting levels of processing. The result of this highly entwined mesh of processes is supreme robustness, efficiency, and versatility. ORGANIC adopts principles of cortical architecture and self-organizing neurodynamics for the design of a new type of cognitive architectures for speech/language tasks. The neurodynamical models will be grounded in the paradigm of Reservoir Computing is a biologically inspired perspective on how arbitrary computations can be learnt and performed in artificial neural networks which are -- like their biological role models -- large, randomly grown, highly nonlinear and eminently adaptive. R\&D activities in Organic will result in: - a much deeper theoretical understanding of how very complex computations, especially those related to language processing, can be robustly and adaptively performed in neurodynamical systems, - a publicly available "Engine'' of programming tools which conforms to recent interface standards for parallel neural system simulations, together with a reference collection of large real-life benchmark datasets, - prototype implementations of large-vocabulary online speech recognizers and handwriting recognition solutions. The consortium brings together the original pioneers in reservoir computing, leading researchers in cortical architectures for speech and language processing, speech recognition technology and an industrial partner at the frontier of automated text recognition.
Start: 31.03.2009
End: 30.03.2012
EU SECO - Self-Constructing Computing Systems
Although the density and speed of integrated circuits has grown exponentially during the last decades, so too have costs of fabrication and test facilities. Current computational methods depend on completely reliable hardware, a constraint that greatly increases the degree of fabrication precision required to avoid failure of individual components, and also increases the amount of post-fabrication testing required to confirm correct function. Nature has solved these production problems. The neocortex, is a cellular computer that generates intelligent behavior. But more than this, it constructs and configures itself by replication of a few precursor cells. Each derived cell is equipped with a set of related rules inherited from its functional parents, and by each cell implementing these locally, the overall cell mass is able to achieve global coherent action. Harnessing these principles for artificial fabrication would revolutionize computer technology. Here we propose some first steps towards understanding these developmental construction mechanisms. We will demonstrate, by a fusion of experimental neuroscience, detailed physical simulation, and theoretical analysis, the principles by which a population of real or artificial neurons can grow and assemble themselves into functioning circuits. We will apply these principles by engineering some first self-constructing applications, in which a designed genetic code is inserted into a precursor cell, and so initiates a developmental process of cell division, cell migration, neurite growth, and synaptogenesis. The final global organization of self-constructed neural networks appears to be an attractor in which component neurons are able to satisfy their own local organizational objectives. Thus, unlike existing artificial processing systems, we expect our self-constructed networks to respond to damage or environmental changes by significant self-repair and axonal re-wiring.
Start: 29.02.2008
End: 28.02.2012
Doctoral Program: Confluence of Vision and Graphics
Computer vision and computer graphics constitute two closely related areas of research: Though both fields rely on the same physical and mathematical principles and on a common set of representations, they mainly differ in how these representations are built. Traditionally these two fields have been treated as separate academic discipline. Exploiting the commonalities between vision and graphics turns out to be a scientifically profitable endeavour. There are many examples of fruitfull combination of graphics and vision, but there is no systematic education of students (especially in Austria). Therefore, the goal of this doctoral program Confluence of Vision and Graphics is to educate highly talented PhD students in this interdisciplinary field and to teach them a common view of this challenging topic from the start. All proposed topics require a significant amount of vision and graphics. The students will be co-supervised jointly by one professor with vision and one professor with graphics expertise. The proposed educational program will ensure that the students will be trained to become future leading scientists, which will face the challenges of research excellence in the interdisciplinary area of graphics and vision, academic leadership, and social competence as a member of a particular research group as well as being a part of the global research network.
Start: 31.05.2007
End: 29.04.2011
FWF - Cognitive Vision - Computermodelle für Biological Vision Systeme
Die Bildverarbeitung bei Affen und anderen Tieren dient als Hauptquelle für neue Ideen und Algorithmen im Bereich des kognitiven Sehens. Dieses Teilprojekt ermöglicht es dem gemeinsamen Forschungsprojekt (FP) wichtige neue Ergebnisse über die Interaktion niedrigerer und höherer visueller Bereiche im Kortex für kognitives Sehen und über die Funktion des Lernens bei verschiedenen Aufgaben des kognitiven Sehens in seine Arbeit einzubinden. Es werden Computermodelle erstellt, die diese neuen experimentellen Ergebnisse in funktionelle Modelle integrieren und dadurch neue Erkenntnisse über die Interaktion von bottom-up und top-down Bildverarbeitung sowie neue Erkenntnisse über die Funktion von Lernen beim Sehen von Primaten liefern. Somit kann die technisch orientierte Forschung an diesem gemeinsamen FP direkt auf experimentelle Ergebnisse im Bereich kognitiven Sehens in biologischen Organismen zurückgreifen. Zusätzlich wird die enge - und auch einzigartige - Zusammenarbeit zwischen Forschern der Computational Neuroscience und des Maschinellen Sehens innerhalb dieses gemeinsamen FP dazu beitragen, dass Forscher aus der Computational Neuroscience ein besseres Verständnis der Rechenprobleme, die es beim kognitiven Sehen zu lösen gilt, erlangen.
Start: 30.11.2003
End: 14.12.2009
EU - FACETS - Fast Analog Computing with Emergent Transient States in Neural Architecture
Information science has been a major driving force of the economical and social development in the 20th century. Based on the ingenious concept of Alain Turings universal computing machine and the availability of semiconductor based transistors, the IT industry has been able to follow an aggressive roadmap of ever increasing performance according to power laws like the well known Moore's Law. It appears to be a matter of time only until computers will eventually reach the capabilities of the human brain. Upon closer inspection, however, the brain is dramatically different from conventional computers. The differences are not only due to the use of biological tissue rather than silicon but also in terms of the computing architecture. The brain is not composed out of highly specialized and separated building blocks like a microprocessor but exhibits a rather uniform structure. It does not use Boolean operations like ANDs and ORs to perform logical operations on well defined stable states but involves the dynamics of transient states to code and to process information. Maybe most importantly, there is no engineered software to deal with pre-defined situations. Instead, the brain is based on a huge number of truly massively parallel non-linear processing elements (neurons), a very high connectivity (synapses) and self-organisation (learning, plasticity). The FACETS project aims to address the unsolved question of how the brain computes with a concerted action of neuroscientists, computer scientists, engineers and physicists. It combines a substantial fraction of the European groups working in the field into a consortium of 13 groups from Austria, France, Germany, Hungary, Sweden, Switzerland and the UK. About 80 scientists join their efforts over a period of 4 years, starting in September 2005. A project of this dimension has rarely been carried out in the context of brain-science related work in Europe, in particular with such a strong interdisciplinary component.
Start: 31.08.2005
End: 30.08.2009
FWF-Adaptve RobotSteurg - Adaptive Control of Humanoid Robots based on Principles of Computation and Learning in Neural Microcircuits
The performance of humanoid robots and other complex robotic systems in natural environments is still unsatisfactory. This interdisciplinary research project develops novel solutions for robot control, especially for humanoid robots, that emerge from new insight into real-time computing and adaptivity in neural microcircuits of biological organisms. This strategy leads to the investigation of novel computational organizations and learning strategies for sensory-motor loops that integrate on a more abstract level salient principles of biological motor control. The research will be carried out jointly by robotic experts from the BIRG (Biologically Inspired Research Group) at EPFL (Lausanne), the Brain Mind Institute at EPFL (Lausanne), and the Institut für Grundlagen der Informationsverarbeitung, TU Graz, with funding provided by the Swiss National Science Foundation, the Brain Mind Institute at EPFL, and the FWF in Austria.
Start: 14.04.2004
End: 30.03.2009
EU - PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning - EU - PASCAL Network of Excelle
PASCAL is a FP6 Network of Excellence within the 6th Framework Programme for research and technological development of the EU. The theme of the network is the confluence of pattern analysis, statistical modelling, computational learning and large-scale optimization to create new and very powerful methods for analyzing large quantities of sensed or warehoused data. Pascal will consolidate this effort, providing a core of top-ranking young researchers with in-depth training in this emerging discipline, and further breaking down interdisciplinary barriers by allowing focused research attacks on key "watershed" issues to be funded. Topics covered will include: computational learning theory for adaptive systems and for problems with many overlapping classes; the relationship between Bayesian and frequentist learning; optimization methods for large-scale statistical modelling and learning problems; advanced kernel based methods for visual object detection, speech analysis, and text/web mining; and learning based approaches to user modelling and data fusion.
Start: 30.11.2003
End: 28.02.2008
Measurable intelligent and reliable semantic extraction and retrieval of multimedia data - MISTRAL
Multimedia data has a rich and complex structure in terms of inter- and intra-document references and can be an extremely valuable source of information. However, this potential is severely limited until and unless effective methods for semantic extraction and semantic-based cross-media exploration and retrieval can be devised. Today’s leading-edge techniques in this area are working well for low-level feature extraction (e.g. colour histograms), are focussing on narrow aspects of isolated collections of multimedia data, and are dealing only with single media types. MISTRAL follows the following lines of radically new research: MISTRAL will extract a large variety of semantically relevant metadata from one media type and integrate it closely with semantic concepts derived from other media types. Eventually, the results from this cross-media semantic integration will also be fed back to the semantic extraction processes of the different media types so as to enhance the quality of the results of these processes. MISTRAL will focus on most innovative, semantic-based cross-media exploration and retrieval techniques employing concepts at different semantic levels. MISTRAL addresses the specifics of multimedia data in the global, networked context employing semantic web technologies. The MISTRAL results for semantic-based multimedia retrieval will contribute to a significant improvement of today’s human-computer interaction in multimedia retrieval and exploration applications. New types of functionalities include but are not limited to *cross-media-based automatic detection of objects in multimedia data: For example, if a video contains an audio stream with barking together with a particular constellation of video features, the system can automatically consider the features in the video as an object “dog”. *semantic-enriched cross-media queries: A sample query could be “find all videos with a barking dog in the background and playing children in the foreground”. *cross-media synchronisation: The idea is to synchronize independent types of media according to the extracted semantic concepts. For example, if users see somebody walking in a video, they should also hear footfall from an audio.
Start: 31.12.2004
End: 29.09.2007
MISTRAL - Measurable Intelligent and Reliable Semantic Extraction and Retrieval of Multimedia Data
Multimedia data has a rich and complex structure in terms of inter- and intra-document references and can be an extremely valuable source of information. However, this potential is severely limited until and unless effective methods for semantic extraction and semantic-based cross-media exploration and retrieval can be devised. Today’s leading-edge techniques in this area are working well for low-level feature extraction (e.g. colour histograms), are focussing on narrow aspects of isolated collections of multimedia data, and are dealing only with single media types. MISTRAL follows the following lines of radically new research: MISTRAL will extract a large variety of semantically relevant metadata from one media type and integrate it closely with semantic concepts derived from other media types. Eventually, the results from this cross-media semantic integration will also be fed back to the semantic extraction processes of the different media types so as to enhance the quality of the results of these processes. MISTRAL will focus on most innovative, semantic-based cross-media exploration and retrieval techniques employing concepts at different semantic levels. MISTRAL addresses the specifics of multimedia data in the global, networked context employing semantic web technologies. The MISTRAL results for semantic-based multimedia retrieval will contribute to a significant improvement of today’s human-computer interaction in multimedia retrieval applications.
Start: 31.12.2003
End: 30.12.2006
Computer Models for Biological Vision Systems
Visual processing in primates and other species serves as a primary source for new ideas and algorithms in cognitive vision. This subproject will allow the JRP to integrate into its work some of the most exciting new experimental data about primate vision, especially new data regarding the interaction of lower and higher visual areas in cortex and the role of learning in various cognitive vision tasks. Computer models will be designed that integrate these new experimental data into functional models, and thereby provide new insight into the interaction of bottom-up and top-down processing in primate vision, as well as new insight into the role of learning principles in primate vision. In this way the engineering-oriented research in this JRP does not have to rely on secondary (and often somewhat dated) reports about experimental findings on cognitive vision in biological organisms, but can go directly to the source. In addition the close - and quite unique - interaction between researchers in computational neuroscience and machine vision within this JRP will help researchers from computational neuroscience to gain a better understanding of the computational problems that cognitive vision has to solve, and will help to sharpen the experimental questions addressed in research on primate vision.
Start: 14.12.2003
End: 13.12.2006
Computing and Learning in Circuits of Spiking Neurons
Information is processed in the neocortex by extremely complex but surprisingly stereotypic circuits of neurons. The goal of this project is to understand the organization of computation and learning in neural microcircuits, which form the lowest level of circuit organization in the cortex. This research will be carried out through theoretical analysis and computer simulation, which take the most recent biological data into account. A unique aspect of this project is the close collaboration of computer scientists and neuroscientists. The team of Prof. Maass at the Institute for Theoretical Computer Science (TU Graz) will collaborate with the team of the neuroscientist Prof. Markram from the Weizmann Institute in Israel, who is one of the leading experts for the experimental investigation of neural microcircuits in the neocortex. The project builds on results from a preceding FWF-project, that had focused on the components of neural circuits: spiking neurons and dynamic synapses. It is expected that the understanding of the organization of information processing in neural circuits will also provide new ideas for the design of novel artificial computing machinery ("neuromorphic engineering"). In this project computer models of neural microcircuits will be used to explore new methods for training a robot to respond in real-time to rapidly changing input.
Start: 31.05.2002
End: 30.12.2005
Special Research Area (SFB) "Optimization and Control"
By combining subprojects which are in part devoted to applied mathematics and in part to concrete applications into a single organizational unit over a longer period of time, the following basic objectives should be accomplished: a) Concentration and further development of the problem-solving capacity in the areas of control and optimization in Graz. b) Intensify the cooperation between applied mathematics and applications as being represented in the Special Research Area. c) Strengthen the application-oriented components in the mathematical education at both universities in Graz. The present constellation of subprojects of this SFB is very appropriate in view of the basic objectives listed above and also provides the necessary spectrum of methods in order to deal with problems of mathematical modeling and simulation. It was possible to intensify already existing cooperations and to establish new ones. The mathematical subprojects are confronted with stimulating problems considered in the applied subprojects, whereas the latter have access to the mathematical know-how of the mathematical subprojects. The results obtained during the first two funding periods are documented in a large number of publications, in the centre report and in the progress reports for the periods 1994-1997 and 1997-2000. The international cooperations of the different groups were considerably intensified since the SFB was initiated. not assigned KP: Institute und Mitarbeiter der Karl-Franzens-Universität Graz
Start: 31.07.1994
End: 30.12.2003
Clustering and Triangulationproblems
Start: 31.12.2000
End: 30.12.2003
Learning for Adaptable - Learning for Adaptable Visual Assistants (LAVA)
The key problem that must be solved in order to build cognitive vision systems is the robust, efficient and learnable categorisation and interpretation of large numbers of objects, scenes and events, in real settings. LAVA will create technologies enabling such systems and an understanding of the systems- and user-level aspects of their applications, via a novel alliance between statistical learning theory, computer vision and cognitive science experts. For practical computational efficiency and robustness, we shall devise methods for goal-directed visual attention and the integration of multiple asynchronous visual cues. These results will be embodied in two integrated systems: one will employ vision for information retrieval in a mobile setting; the other will derive symbolic representations from video sequences, enabling a wide range of "ambient intelligence scenarios.
Start: 30.04.2002
End: 30.12.2003
Neural and Computational Learning II - NeuroCOLT II (EU)
Start: 31.05.1998
End: 30.05.2002
Computing and Learning with Spiking Neurons (FWF)
This Research Project compared traditional models for "artificial neural networks" with biologically more realistic models for neurons and synapses. There exist significant differences between these two types of models, especially since biological neurons encode their output in the form of brief electrical pulses ("spikes"), and since biological synapses change their "weight" on a very fast time scale in dependence of the spike pattern that has previously reached them. This Research Project developed models for information processing with these more realistic types of neurons and synapses, thereby building a foundation for a new generation of neural networks models. Most of the research results form this project are online available (http://www.igi.TUGraz.at/maass/publications.html, and publications on http://www.igi.tugraz.at/tnatschl/, http://www.igi.tugraz.at/legi/). In addition some of them were also described in popular science articles (http://www.igi.tugraz.at/maass/nonexperts.html).
Start: 31.03.1997
End: 30.03.2002
Publications
Start: 31.12.1994
End: 30.01.2002
Navigation of an Autonomous Mobile Robot Supported by Supplementary Information
Start: 31.12.1994
End: 30.01.2002
Triangulations and Surface Modeling (OeAD)
Surface modelling using triangular irregular networks (TINs) is an important area within geometric data processing, with applications to medical images, computer graphics, and finite elements methods. The project aims at improving existing triangulation methods, especially for non-convex surfaces. Appropriate generalizations of Delaunay triangulations and greedy triangulations for such surfaces are investigated.
Start: 31.12.1994
End: 30.07.2001
Associate Note of NeuroNet (Network of Excellence in Neural Networks), EU
Start: 31.12.1994
End: 30.05.2001
Intelligent Machines
Start: 31.12.1998
End: 30.12.2000
Learnable Mobile Robots
Start: 14.03.2000
End: 30.10.2000
Theory of Machine Learning and Neural Networks (BMWF)
Start: 31.12.1994
End: 30.01.2000
Algorithms on Hypercubes - A Geometric Approach to String Searching and Coding (BMWF)
The aim of the project is to investigate geometric and combinatorial properties of the so-called hypercube (generalization of the usual 3D cube to d dimensions). Emphasis is laid on the intersection of the hypercube with various hyperplanes. Since the vertices of the hypercube constitute a compact representation of all possible binary strings of length d, applications of the results arise in coding theory and binary string searching.
Start: 31.12.1994
End: 30.12.1998
Neural and Computational Learning I - NeuroCOLT I (EU)
Start: 31.03.1995
End: 30.03.1998
Algorithms on Hypercubes (OeNB)
The aim of the project is to investigate geometric and combinatorial properties of the so-called hypercube (generalization of the usual 3D cube to d dimensions). Emphasis is laid on the intersection of the hypercube with various hyperplanes. Since the vertices of the hypercube constitute a compact representation of all possible binary strings of length d, applications of the results arise in coding theory and binary string searching.
Start: 29.10.2018
End: 30.07.1995

Contact
image/svg+xml

Institute of Theoretical Computer Science
Inffeldgasse 16b/I
8010 Graz

Phone: +43 (0) 316 / 873 - 5811
Fax:       +43 (0) 316 / 873 - 105811
daniela.windisch-scharlernoSpam@tugraz.at


Head
Assoc. Prof. Dipl.-Ing. Dr. techn. Robert Legenstein