BibSonomy publications for /user/flint63/analysisMon Jul 07 13:47:29 CEST 2008Berlin, HeidelbergLecture Notes in Computer ScienceMultimodal Technologies for Perception of Humans: First International Evaluation Workshop on Classification of Events, Activities and Relationships, CLEAR 2006, Southampton, UK, April 6-7, 2006, Revised Selected Papers41222007pattern book action user analysis sensor traffic interaction data video image ai speech multimodal recognition v0805 This book constitutes the thoroughly refereed post-proceedings of the First International CLEAR 2006 Evaluation Campaign and Workshop on Classification of Events, Activities and Relationships for evaluation of multimodal technologies for the perception of humans, their activities and interactions, held in Southampton, UK, in April 2006. The 29 revised full system description papers and 1 institutional paper presented together with 2 invited papers were carefully reviewed and selected for inclusion in the book. CLEAR is an international effort to evaluate systems that are designed to analyze people's identities, activities, interactions and relationships in human-human interaction scenarios, as well as related scenarios. The papers are organized in topical sections on 3D person tracking, 2D face detection and tracking, person tracking on surveillance data, vehicle tracking, person identification, head pose estimation, acoustic scene analysis, and other evaluations.Mon Jul 07 13:47:29 CEST 2008Berlin, Heidelberg{Fahrerassistenzsysteme mit maschineller Wahrnehmung}2005assist interaction information interface analysis ai user v0805 data design engineering traffic embedded book management sensor processing Fahrerassistenzsysteme unterstützen den Fahrer in seiner Fahraufgabe und entlasten ihn dadurch gezielt. Viele Experten erwarten, dass Fahrerassistenzsysteme zur Sicherheit des Straßenverkehrs wesentlich beitragen werden. Zentrale wissenschaftliche und industrielle Herausforderungen bestehen zur Zeit in der Erforschung und Entwicklung maschineller Wahrnehmungsfähigkeiten, die eine angemessene Erfassung der Umwelt und deren fahrergerechte Integration in geeignete Fahrfunktionen leisten. Dieser Band basiert auf ausgewählten Vorträgen eines Workshops in Walting (Altmühltal) und macht deren Inhalt in erweiterter Fassung zugänglich. In bislang nicht vorliegender Interdisziplinarität diskutieren Experten aus Wissenschaft und Praxis unterschiedlichste Ansätze aus vielfältigen Bereichen wie der maschinellen Wahrnehmung, Mensch-Maschine-Interaktion, Wissensrepräsentation, Funktionsentwicklung und Wirtschaftsethik. Über die fachlich-technische Auseinandersetzung mit Fahrerassistenzsystemen hinaus wird damit auch ein Beitrag zum notwendigen Diskurs über deren Auswirkung und gesellschaftliche Akzeptanz geleistet.Mon Jul 07 13:47:29 CEST 2008AmsterdamFrontiers in Artificial Intelligence and ApplicationsOntology Learning from Text: Methods, Evaluation and Applications1232005processing text v0805 ai book semantic ontology language web learn analysis This volume brings together ontology learning, knowledge acquisition and other related topics. It presents current research in ontology learning, addressing three perspectives. The first perspective looks at methodologies that have been proposed to automatically extract information from texts and to give a structured organization to such knowledge, including approaches based on machine learning techniques. Then there are evaluation methods for ontology learning, aiming at defining procedures and metrics for a quantitative evaluation of the ontology learning task; and finally application scenarios that make ontology learning a challenging area in the context of real applications such as bio-informatics. According to the three perspectives mentioned above, the book is divided into three sections, each including a selection of papers addressing respectively the methods, the applications and the evaluation of ontology learning approaches.Mon Jul 07 13:47:29 CEST 2008AmsterdamFrontiers in Artificial Intelligence and ApplicationsOntology Learning and Population: Bridging the Gap between Text and Knowledge1672008processing text analysis v0805 semantic ontology learn book language web ai The promise of the Semantic Web is that future web pages will be annotated not only with bright colors and fancy fonts as they are now, but with annotation extracted from large domain ontologies that specify, to a computer in a way that it can exploit, what information is contained on the given web page. The presence of this information will allow software agents to examine pages and to make decisions about content as humans are able to do now. The classic method of building an ontology is to gather a committee of experts in the domain to be modeled by the ontology, and to have this committee agree on which concepts cover the domain, on which terms describe which concepts, on what relations exist between each concept and what the possible attributes of each concept are. All ontology learning systems begin with an ontology structure, which may just be an empty logical structure, and a collection of texts in the domain to be modeled. An ontology learning system can be seen as an interplay between three things: an existing ontology, a collection of texts, and lexical syntactic patterns. The Semantic Web will only be a reality if we can create structured, unambiguous ontologies that model domain knowledge that computers can handle. The creation of vast arrays of such ontologies, to be used to mark-up web pages for the Semantic Web, can only be accomplished by computer tools that can extract and build large parts of these ontologies automatically. This book provides the state-of-art of many automatic extraction and modeling techniques for ontology building. The maturation of these techniques will lead to the creation of the Semantic Web.Mon Jul 07 13:47:29 CEST 2008it -- Information Technology117-24{Mit aktiven Sensoren das Kfz-Umfeld erfassen -- Funktion und Leistungsfähigkeit von Radar \& Co}492007paper information v0805 processing traffic analysis sensor data recognition ai Aktive Umfeldsensoren senden aktiv Strahlung aus. Die Rückstreuung dieser Strahlung ermöglicht die Erkennung von Umfeldobjekten und die Messung von Abstand und Raumlage. Die spezifischen Funktionsweisen der im Kraftfahrzeugbereich eingesetzten Sensorprinzipien mit Ultraschall, Licht- und mm-Wellen werden in diesem Beitrag beleuchtet. Die Weiterverarbeitung der Signale und Konzepte für eine Sensordatenfusion werden vorgestellt. Abschließend werden einige der nichtfunktionalen, aber oftmals erfolgsentscheidenden Kriterien für den Einsatz im Automobil angesprochen. Detecting the Automotive Surrounding with Active Sensors -- Function and Capabilities of Radar & Co.: Active surroundings sensors emit rays actively. Back-scattering these rays enables the detection of surrounding objects and measurement of range and position. The specific kind of function of the current automotive surrounding sensors are illustrated in the course of this paper. The basics of processing and concepts for sensor data fusion are presented. Some of the non-functional criteria are discussed at the end. They are prerequisites for a successful equipment in automobilesMon Jul 07 13:47:29 CEST 2008it -- Information Technology13-4Fahrerassistenzsysteme492007recognition paper cognitive traffic analysis assist sensor generation data ai v0805 action Besonders der Automobilsektor bildet ein viel beachtetes Anwendungsfeld für die Forschung im Bereich maschineller Kognition. So genannte Fahrerassistenzsysteme nehmen die Fahrzeugumgebung sensoriell wahr und generieren aus der Interpretation der aktuellen Situation heraus angemessenes automatisches Handeln. Editorial for Special Issue on Driver Assistance SystemsMon Jul 07 13:47:29 CEST 2008Berlin, HeidelbergArtifical Intelligence for Human Computing: {ICMI} 2006 and {IJCAI} 2007 International Workshops47-71Lecture Notes in Artificial IntelligenceHuman Computing and Machine Understanding of Human Behavior: A Survey44512007recognition ai interface paper multimodal v0805 user springer analysis A widely accepted prediction is that computing will move to the background, weaving itself into the fabric of our everyday living spaces and projecting the human user into the foreground. If this prediction is to come true, then next generation computing should be about anticipatory user interfaces that should be human-centered, built for humans based on human models. They should transcend the traditional keyboard and mouse to include natural, human-like interactive functions including understanding and emulating certain human behaviors such as affecti0ve and social signaling. This article discusses how far are we from enabling computers to understand human behavior.Mon Jul 07 13:47:29 CEST 2008it -- Information Technology133-{Videobasierte 4D-Umfelderfassung für erweiterte Assistenzfunktionen}492007sensor traffic v0805 image ai video paper information analysis assist processing Mit dem verstärkten Einzug von Assistenzfunktionen ins Fahrzeug steigen auch die Anforderungen an eine maschinelle Umfelderfassung. Während sich bislang großteils auf Radarsensorik konzentriert wurde, rechtfertigen Funktionen wie beispielsweise ein erweitertes ACC oder STA zunehmend auch den Einsatz von Videosensorik. In diesem Kontext wird der so genannte 4D-Ansatz zur Szeneninterpretation, eine daraus abgeleitete allgemeine Systemstruktur sowie ein Satz grundlegender Bildverarbeitungsoperatoren vorgestellt. Aufbauend darauf werden als konkrete Anwendungen eine videobasierte Fahrspur- und Fahrzeugerkennung präsentiert sowie deren kombinierte Nutzung im Rahmen einer Demonstration zum Thema Stauassistent. Vision-Based 4D-Environmental Perception for Advanced Assistance Functions: The demand for improved driver assistance functionality is increasing, thus, the requirements for automatic environmental perception has also increased. Previous efforts have concentrated on radar sensors, however advanced ACC or STA applications would benefit from fusion with vision in the future. In this article the 4D-approach to scene interpretation is introduced, and a general system architecture derived from it including a basic set of image processing operations is proposed. An example of a vision-based lane and vehicle tracking system is presented and finally demonstrated within an STA application.Mon Jul 07 13:47:29 CEST 2008Ume{\r{a}}, SwedenFace Recognition: A Single View Based {HMM} Approach2008user image book recognition analysis v0805 ai This dissertation addresses the challenges of giving computers the ability of doing face recognition, i.e. discriminate between different faces. Face recognition systems are commonly trained with a database of face images, becoming “familiar” with the given faces. Many reported methods rely heavily on training database size and representativenes. But collecting training images covering, for instance, a wide range of viewpoints, different expressions and illumination conditions is difficult and costly. Moreover, there may be only one face image per person at low image resolution or quality. In these situations, face recognition techniques usually suffer serious performance drop. Here we present effective algorithms that deal with single image per person database, despite issues with illumination, face expression and pose variation. Illumination changes the appearance of a face in images. Thus, we use a new pyramid based fusion method for face recognition under arbitrary unknown lighting. This extended approach with logarithmic transform works efficiently with a single image. The produced image has better contrast at both low and high ranges, i.e. has more visible details than the original one. An improved method works with high dynamic range images, useful for outdoor face images. Face expressions also modify the images’ appearance. An extended Hidden Markov Models (HMM) with a flexible encoding scheme treats images as an ensemble of horizontal and vertical strips. Each person is modeled by Joint Multiple Hidden Markov Models (JM-HMMs). This approach offers computational advantages and the good learning ability from just a single sample per class. A fast method simulated JM-HMM functionality is then derived. The new method with abstract observations and a simplified similarity measurement does not require retraining HMMs for new images or subjects. Pose invariant recognition from a single sample image per person was overcome by using the wire frame Candide face model for the synthesis of virtual views. This is one of the support functions of our face recognition system, WAWO. The extensive experiments clearly show that WAWO outperforms the state-of-the-art systems in FERET tests.Mon Jul 07 13:47:29 CEST 2008Computer452--59Analysis and Semantic Querying in Large Biomedical Image Datasets412008retrieval information v0805 spatial ai grid ieee semantic paper analysis rdf image health Biomedical image analysis plays an important role in diagnosing, prognosing, and treating complex diseases. The authors describe a set of techniques for analyzing, processing, and querying large image datasets using semantic and spatial information.Mon Jul 07 13:47:29 CEST 2008New York, USA{TMR '07:} Proceedings of the 2007 Workshop on Tagging, Mining and Retrieval of Human Related Activity Information35-42Tagging Strategies for Extracting Real-World Events with Networked Sensors2007sensor recognition analysis ai action v0805 language information temporal paper acm embedded In this paper, we introduce our 's-room' project as well as the tagging strategies and environment developed for the project. In the s-room, many small sensor nodes are attached to various objects. Our project aims to construct a system for comprehending real-world events and the properties or status information of physical objects by utilizing sensor nodes distributed throughout the room as well as general knowledge obtained from web space. The events extracted in the s-room are then published as web contents. We defined a set of event descriptors as a middle language between the sensor data stream and natural language description. The descriptors are selected by a two-way method: 1) a top-down approach based on definitions in NL-dictionaries and laws in physics, 2) a bottom-up approach based on manually tagged sensor data streams. We also developed a tagging environment that enables us to arrange the relationship between NL phrase expressions of human activities and multiple sensor events automatically extracted from the sensor signal streams.Mon Jul 07 13:47:29 CEST 2008New York, USA{TMR '07:} Proceedings of the 2007 Workshop on Tagging, Mining and Retrieval of Human Related Activity Information27-34Modeling Temporal Dependencies between Observed Activities2007knowledge ai acm analysis v0805 temporal processing recognition paper concurrent The modeling of parallel activities requires a notation which can represent the temporal dependencies as well as variations of the execution order of the activities. This paper introduces ART (Activity Relation Trees), a notation to describe temporal dependencies between activities. ART is based on ConcurTaskTrees (CTT) that are extended with the means to describe temporal relationships. Furthermore, we present an algorithm that allows to automatically generate ART models from observed examples. Because former approaches for automatic model acquisition were restricted to strictly sequential data and cannot be applied in the case of parallel activities, we developed a method to reduce the problem of automatic modeling of parallel activities to the simpler task of modeling sequential data. By grouping activities and distinguishing different phases we are able to form general descriptions of a scenario that include variations in the execution order. The paper defines all necessary concepts and describes the algorithm in detail. The evaluation of the algorithm shows that precise models can be generated by using only few examples.Mon Jul 07 13:47:29 CEST 2008it -- Information Technology15-16Kreuzungsverstehen -- Ein wissensbasierter Ansatz492007analysis image traffic map processing ai knowledge recognition data paper v0805 sensor location video Existierende Ansätze zur Fahrbahnerkennung arbeiten -- vermutlich prinzipbedingt -- auf jeweils nur kleinen Teilmengen der möglichen Fahrbahnkonfigurationen robust. Ein Schlüssel auf dem Weg zu allgemeingültig arbeitenden Verfahren liegt in der massiven Erhöhung der Menge des im Schätzprozess genutzten Vorwissens. Im Hinblick auf eine effiziente, nachvollzieh- und erweiterbare Beschreibung ist eine explizite Repräsentation solchen Wissens anzustreben. Im vorliegenden Beitrag wird eine konzeptuelle und eine geometrische Wissensrepräsentationen für den Diskursbereich Straßen und Kreuzungen vorgestellt. Deren Parameter werden gemeinsam mittels eines Multihypothesen-Ansatzes geschätzt, dessen Eingangsdaten handelsübliche digitale Karten sowie verschiedene videobasierte Objektdetektoren sind. Zur Verifikation der Hypothesen werden die Vorzugsrichtungen der lokalen Textur im Bereich der erwarteten Fahrspurberandungen ausgewertet. Dabei kann gleichzeitig die Messunsicherheit der zur Bildprojektion genutzten Schätzung der Lage des Kamerakoordinatensystems reduziert werden. Junction Understanding---A Knowledge Based Approach: The road recognition problem has been solved robustly only for small, often simplified subsets of possible road configurations. A massive augmentation of the amount of used prior knowledge could pave the way towards generally valid estimators. An explicit representation of such knowledge will additionally lead to an efficient, understandable and therefore extendible system. We present a conceptual and a geometrical knowledge representation for the Roads&Junctions domain of discourse. Its parameters are estimated using a multi hypotheses approach. A commercially available digital map and a set of video based object detectors serve as input data. The resulting hypotheses are verified by evaluating the preferred orientations of local texture around the expected position of the lane dividers. The estimate of the camera coordinate system's pose, which is used for image projection, is updated simultaneously.Mon Jul 07 13:47:29 CEST 2008Informatik-Spektrum13-14{Architektur von Data Warehouses und Business Intelligence Systemen}282005analysis data business v0805 springer paper recognition database pattern interface information enterprise Business Intelligence (BI) ist der Prozess der Umwandlung von Daten in Informationen und weiter in Wissen. Entscheidungen und Prognosen stützen sich auf dieses Wissen und schaffen dadurch Mehrwert für ein Unternehmen. Ein Data Warehouse (DW) bildet in vielen Fällen die technische Basis zur Implementierung einer BI-Lösung.Mon Jul 07 13:47:29 CEST 2008CME 2007: IEEE/ICME International Conference on Complex Medical Engineering416-420Wireless {MEMS} Sensing System for Human Activity Monitoring2007recognition sensor check v0805 datamining information analysis ieee ai embedded paper temporal This study describes a human activity monitoring system that uses microelectromechanical systems (MEMS) sensors. The prototype system contains four sensors for ambient monitoring: 3-axis acceleration, barometric pressure, temperature, and relative humidity. The peripheral circuitry for each sensor is connected to a one-chip microprocessor. The measured data is stored in memory and transferred via a wireless transmitter. We measured a human subject's daily life activities, and using data mining, we were able to obtain a representation of that subject's life circumstances.Mon Jul 07 13:47:29 CEST 2008it -- Information Technology125-32Kollisionsvermeidung durch raum-zeitliche Bildanalyse492007sensor recognition paper action analysis v0805 assist information video ai processing traffic image Mehr als 1/3 aller Unfälle mit Personenschäden passieren im städtischen Bereich, primär an Kreuzungen. Eine Unterstützung des Fahrers durch geeignete Assistenzsysteme erfordert das Verstehen dieser sehr komplexen Situationen, insbesondere das sichere Erkennen anderer bewegter Verkehrsteilnehmer. Der Beitrag zeigt, wie man durch eine geschickte Fusion von Stereosehen und Bewegungswahrnehmung zu einer robusten und schnellen Detektion relevanter bewegter Objekte kommt. Dabei schätzt das als 6D-Vision bezeichnete Verfahren simultan Ort und Bewegung einzelner Bildpunkte und erlaubt somit eine Detektion bewegter Objekte bereits auf Pixelebene. Unter Verwendung eines Kalman-Filters propagiert der Algorithmus die aktuelle Interpretation ins nächste Bild, sodass er sich in Echtzeit darstellen lässt. Beispiele kritischer Situationen im Innenstadtbereich verdeutlichen die Leistungsfähigkeit des 6D-Vision-Prinzips, das auch im Bereich der mobilen Roboter wertvolle Beiträge leisten kann. Collision Avoidance based on Space-Time Image Analysis: More than one third of all traffic accidents with injuries occur in urban areas, especially at intersections. A suitable driver assistance system for such complex situations requires the understanding of the scene, in particular a reliable detection of other moving traffic participants. This contribution shows how a robust and fast detection of relevant moving objects is obtained by a smart combination of stereo vision and motion analysis. This approach, called 6D Vision, estimates location and motion of pixels simultaneously which enables a detection of moving objects on a pixel level. Using a Kalman Filter, the algorithm propagates the current interpretation to the next image. Hence a real-time implementation is achieved. Examples of critical situations in urban areas exhibit the potential of the 6D Vision concept which can also be extended to robotics applications.Mon Jul 07 13:47:29 CEST 2008Berlin, Heidelberg{KI 2007:} Advances in Artificial Intelligence279-292Lecture Notes in Computer ScienceNatural Language Descriptions of Human Behavior from Video Sequences46672007v0805 image analysis paper ai language generation action recognition video springer This contribution addresses the generation of textual descriptions in several natural languages for evaluation of human behavior in video sequences. The problem is tackled by converting geometrical information extracted from videos of the scenario into predicates in fuzzy logic formalism, which facilitates the internal representations of the conceptual data and allows the temporal analysis of situations in a deterministic fashion, by means of Situation Graph Trees (SGTs). The results of the analysis are stored in structures proposed by the Discourse Representation Theory (DRT), which facilitate a subsequent generation of natural language text. This set of tools has been proved to be perfectly suitable for the specified purpose.Mon Jul 07 13:47:29 CEST 2008Berlin, Heidelberg{SmartKom}: Foundations of Multimodal Dialogue Systems195-207Natural Language Understanding2006processing analysis ai v0805 paper smartkom language dfki This chapter presents SPIN, a newly developed template-based semantic parser used for the task of natural language understanding in SmartKom. The most outstanding feature of the approach is a powerful template language to provide easy creation and maintenance of the templates and flexible processing. Nevertheless, to achieve fast processing, the templates are applied in a sequential order that is determined offline.Mon Jul 07 13:47:29 CEST 2008Proceedings of the International Conference on Visual Information Engineering {(VIE 2006)}261-266Adding Semantic Metadata to Audio-Video Material by Automatic Analysis of Complementary Sources2006ai multimedia analysis language semantic v0805 image knowledge smartweb data processing paper video We present in this paper actual work on adding semantic metadata to multimedia material, on the base of the results of the automatic analysis applied to associated language material, being speech transcripts or various types of textual documents related to video/image material.Mon Jul 07 13:47:29 CEST 2008Berlin, Heidelberg{KI 2007:} Advances in Artificial Intelligence19-42Lecture Notes in Computer ScienceCognitive Technical Systems --- What Is the Role of Artificial Intelligence?46672007springer traffic recognition video plan v0805 robot image paper sensor action factory pattern analysis embedded ai data The newly established cluster of excellence CoTeSys investigates the realization of cognitive capabilities such as perception, learning, reasoning, planning, and execution for technical systems including humanoid robots, flexible manufacturing systems, and autonomous vehicles. In this paper we describe cognitive technical systems using a sensor-equipped kitchen with a robotic assistant as an example. We will particularly consider the role of Artificial Intelligence in the research enterprise. Key research foci of Artificial Intelligence research in CoTeSys include (a) symbolic representations grounded in perception and action, (b) first-order probabilistic representations of actions, objects, and situations, (c) reasoning about objects and situations in the context of everyday manipulation tasks, and (d) the representation and revision of robot plans for everyday activity.