EU FP6 Network of Excellence on Multimedia Understanding through Semantics, Computation, and Learning:The MUSCLE network, started in March 2004, aims at creating and supporting a pan-European Network of Excellence to foster close collaboration between research groups in multimedia datamining on the one hand and machine learning on the other. The focus of our contribution lies particulalrly in the field of Single Modality Processing, particulalry in the fields of text and audio analysis.

Video matting is the problem of estimating foreground and background colors as well as an opacity factor for every pixel in an image sequence. Video object segmentation and matting are fundamentally important operations in many image/video editing and compositing tasks, with a number of applications for instance in the entertainment industry. Once a video object has been extracted successfully from its background by appropriate matting techniques, it may be inserted into another video scene and/or combined with artificial objects to create mixed reality applications that merge the aspect of real with computer-generated worlds.The matting problem is severely ill-posed and generally cannot be solved without any prior knowledge. Early methods constrain the problem by filming the object in front of a constant colored background, which is known as blue screen matting. Although this technique has been effectively used in thefilm industry, it cannot be applied to natural images or video sequences. Subsequently a number of approaches have been proposed that extract the foreground object directly from natural scenes. In general these approaches rely on a so-called trimap, which is a segmentation of the image into definitely foreground, definitely background and an unknown region. Fractional opacity values are then computed only for pixels in the unknown region. The manual creation of a good trimap requires a considerable amount of user interaction, in particular for foreground objects with lots of holes and for images with a large amount of semi-transparent foreground. Recently, a matting system was proposed that iteratively computes an opacity value for every pixel of an image based on a small sample of foreground and background pixels marked by the user. This method creates impressive results for a varietyof images without the use of a trimap. However, the iterative matting system leads to a significant increase in computation time and may also get stuck in a local minimum. Although the aforementioned advancements in matting have lead to increasingly good results there are still a lot of problems totackle. Especially the extraction of foreground objects from videos, as opposed to still images, represents a very challenging topic. State-of-the-art video matting systems require a much higher amount of user interaction than for still images and exceed the power of today¿s general purpose computer hardware. In this PhD project, we will explicitly tackle those problems in video matting by developing new and efficient algorithms to quickly extract high-quality video objects from natural image sequences. We seek to develop a new level of advanced user interaction techniques based on eye-tracking technology and parallelization. This will allow us to design interactive matting approaches with iterative improvement of the matte by switching between user interaction and computing intermediate results at high processing rates. The user input will be supported by novel eye-tracking technologyin a multimodal interface design.

Information creation, maintenance and delivery are the primary business processes of institutional tourism intermediaries, which face information heterogeneities in various dimensions. It requires manual coordination tasks, and suffers from missing consensus on agreed concepts and technologies. OnTourism is aimed at (1) applying, concretizing and evaluating Semantic Web technologies such as ontologies, semantic annotation of content, and semantic search to this information-rich and economically quite important domain, (2) identifying, developing and integrating reference ontologies for the tourism industry, and (3) showing the proof-of-concept in a real-world scenario of the Austrian tourism industry. OnTourism seeks to explore and expand the potential of advanced semantic technologies in a research domain of strong national interest, strength, and tradition.

development of open e-learning contents for the vocational learning platform openDock

Measuring is a key to scientific progress. This is particularly true for research concerning complex systems,whether natural or human-built. Multilingual and multimedia information systems are increasingly complex: theyneed to satisfy diverse user needs and support challenging tasks. Their development calls for proper evaluationmethodologies to ensure that they meet the expected user requirements and provide the desired effectiveness.Large-scale worldwide experimental evaluations provide fundamental contributions to the advancement ofstate-of-the-art techniques through common evaluation procedures, regular and systematic evaluation cycles,comparison and benchmarking of the adopted approaches, and spreading of knowledge. In the process, vastamounts of experimental data are generated that beg for analysis tools to enable interpretation and therebyfacilitate scientific and technological progress.PROMISE will provide a virtual laboratory for conducting participative research and experimentation to carry out,advance and bring automation into the evaluation and benchmarking of such complex information systems, byfacilitating management and offering access, curation, preservation, re-use, analysis, visualization, and mining ofthe collected experimental data. PROMISE will: foster the adoption of regular experimental evaluation activities;bring automation into the experimental evaluation process; promote collaboration and re-use over the acquiredknowledge-base; stimulate knowledge transfer and uptake.Europe is unique: a powerful economic community that politically and culturally strives for equality in itslanguages and an appreciation of diversity in its citizens. New internet paradigms are continually extendingthe media and the task where multiple language based interaction must be supported. PROMISE will direct aworld-wide research community to track these changes and deliver solutions so that Europe can achieve one ofits most cherished goals.

Evaluation and Comparison of NLP Tools.Benchmark-based Evaluation of a set of Named-Entity Recognition Tools with respect to qualitative performance and throughput.

The objective of PLANETS is to develop systems and tools which will support the accessibility and use of digital cultural and scientific resources, specifically through the development of novel concepts, techniques and tools to preserve the availability of digital resources over time.

The aim of project ProFiTex is to support fire fighters in their perilous work with a system that supplies mission-relevant information without overwhelming the fire fighter. The design approach will be user-centered with tests starting at an earlypoint in the project to gain maximum user acceptance. Professional fire fighters will be involved from the beginning of the project to ensure, that the system will be tailored to their needs. Project ProFiTex will continue the work of the successfulEU-funded project wearIT@work. The ProFiTex system comprises electronic devices like an infrared camera, localisation sensors and a human-computer interface device integrated into the fire fighters jacket. Since wireless communication is difficult over long distances and through several walls of a building, an innovative method to transmit information will be applied. A security rope carried by fire fighters during a mission shall be equipped with data transmission capabilities. This allows information to be sent outside to the command post and back to the fire fighter. By monitoring several parameters of the fire fighters condition like his movement pattern and stance, problems can be detected immediately. The fire fighter himself is supplied with the possibility to navigate even in smoky environments thanks to the infrared camera and the positioning system implemented into his equipment. Localized information (e.g.¿door¿, ¿victim¿) can be fed into the system using the garment-integrated human-computer interface device. Information will be displayed to the fire fighters, their group leaders and the commander outside the building. The amount and type of information supplid will be carefully chosen, considering the physical danger and psychic stress fire fighters are opposed to. Work safety of fire fighters shall be increased, thus lowering the number of work-related accidents and casualties. Fire fighting missions will be more efficient using the system.

e-Health by definition requires the sharing of patient related datawhen and where necessary. Electronic health records (EHR) promise toimprove communication between health care providers, thus leading tobetter quality of patients'treatment and reduced costs. As highlysensitive patient information provides a promising goal forattackers and is also demanded by insurance companies and employers,there is an increasing social and political pressure regarding theprevention of health data misuse. Within this project we develop ---based on our patent --- the new system PIPE (Pseudonymization ofInformation for Privacy in e-Health). It differs from existingapproaches in its ability to securely integrate primary andsecondary usage of health data. The aim of this project is (a) tobroaden our patented approach to support (semi-)structured metadata, (b) to develop alternative secure storage and retrievaltechniques, (c) to provide for secure storage and access to medicalemergency data, (d) to provide a secure viewer that preventsman-in-the-middle attacks, and (e) to demonstrate our system in thecontext of genome analysis, storage, and retrieval.

E-health requires the sharing of patient-related data when and where necessary. Electronic health records (EHR) promise to improve communication between health care providers, thus improving the quality of patients'treatment and reducing costs. EHRs allow the structured and expandable collection of medical data needed for clinical research studies and thereby not only enable the optimization of clinical studies, but also results in higher statistical significance due to a larger number of samples. While the digitization of medical data and the organization of this data within EHRs have been introduced in some areas, massive amounts of paper-based health records are still produced on a daily basis. This data has to be stored for 30 years for legal reasons but is of no benefit for research organizations, as the unstructured medical data in paper-based health records cannot be efficiently used for clinical studies. Furthermore, legal regulations prohibit the use of documents containing both personal and medical data for clinical studies, which leads to expensive data acquisition phases and limited samples.This project develops a system for the recognition and pseudonymization of personal data in paper-based health records with the overall goal to (i) provide clinical studies with medical data gained from existing paper-based health records while (ii) ensuring patients'privacy by pseudonymizing the (digitized) data. The product integrates unique methods for (i) automatically identifying personal and medical data, (ii) automatically annotating the optical character recognition (OCR) output data of paper-based health records with standard-compliant metadata, and (iii) automatically pseudonymizing the personal data. With the project results, (Austrian) health care organizations profit by (i) strengthen clinical research resulting in faster and more reliable results and reduced costs, and (ii) providing an environment of trust for its patients and employees that guarantees privacy.