Computer analysis of images and patterns is a scienti c eld of longstanding tradition, with roots in the early years of the computer era when electronic brains inspired scientists. Moreover, the design of vision machines is a part of humanity s dream of the arti cial person. I remember the 2nd CAIP, held in Wismar in 1987. Lectures were read in German, English and Russian, and proceedings were also only partially written in English. The conference took place under a di erent political system and proved that ideas are independent of political walls. A few years later the Berlin Wall collapsed, and Professors Sommer and Klette proposed a new formula for the CAIP: let it be held in Central and Eastern Europe every second year. There was a sense of solidarity with scienti c communities in those countries that found themselves in a state of transition to a new economy. A well-implemented idea resulted in a chain of successful events in Dresden (1991), Budapest (1993), Prague (1995), Kiel (1997), and Ljubljana (1999). This year the conference was welcomed at Warsaw. There are three invited lectures and about 90 contributions written by more than 200 authors from 27 countries. Besides Poland (60 authors), the largest representation comes from France (23), followed by England (16), Czech Republic (11), Spain (10), G- many (9), and Belarus (9). Regrettably, in spite of free registration fees and free accommodation for authors from former Soviet Union countries, we received only one accepted paper from Russia."

Driven by the requirements of a large number of practical and commercially - portant applications, the last decade has witnessed considerable advances in p- tern recognition. Better understanding of the design issues and new paradigms, such as the Support Vector Machine, have contributed to the development of - proved methods of pattern classi cation. However, while any performance gains are welcome, and often extremely signi cant from the practical point of view, it is increasingly more challenging to reach the point of perfection as de ned by the theoretical optimality of decision making in a given decision framework. The asymptoticity of gains that can be made for a single classi er is a re?- tion of the fact that any particular design, regardless of how good it is, simply provides just one estimate of the optimal decision rule. This observation has motivated the recent interest in Multiple Classi er Systems , which aim to make use of several designs jointly to obtain a better estimate of the optimal decision boundary and thus improve the system performance. This volume contains the proceedings of the international workshop on Multiple Classi er Systems held at Robinson College, Cambridge, United Kingdom (July 2{4, 2001), which was organized to provide a forum for researchers in this subject area to exchange views and report their latest results.

This book constitutes the refereed proceedings of the Third International Conference on Scale-Space and Morphology in Computer Vision, Scale-Space 2001, held in Vancouver, Canada in July 2001.The 18 revised full papers presented together with 23 posters were carefully reviewed and selected from 60 submissions. The book addresses all current aspects of scale-space and morphology in the context of computer vision, in particular, vector distance functions, optic flow, image registration, curve evolution, morphological segmentation, scalar images, vector images, automatic scale selection, geometric diffusion, diffusion filtering, image filtering, inverse problems, active contours, etc.

This book constitutes the thoroughly refereed post-proceedings of the 10th International Workshop on Theoretical Foundations of Computer Vision, held at Dagstuhl Castle, Germany, in March 2000.
The 20 revised full papers presented have been through two rounds of reviewing, selection, and revision and give a representative assessment of the foundational issues in multiple-image processing. The papers are organized in topical sections on 3D data acquisition and sensor design, multi-image analysis, data fusion in 3D scene description, and applied 3D vision and virtual reality.

This book constitutes the refereed proceedings of the Second International Conference on Advances in Pattern Recognition, ICAPR 2001, held in Rio de Janeiro, Brazil in March 2001.The 40 revised full papers presented together with three invited papers and two tutorial presentations were carefully reviewed and selected for inclusion in the proceedings. The book is organized in topical sections on neural networks and computational intelligence, character recognition and document analysis, feature selection and analysis, pattern recognition and classification, image and signal processing applications, and image feature analysis and retrieval.

The Self-Organising Map (SOM), with its related techniques, is one of the most popular and powerful concepts for unsupervised pattern recognition and data visualisation. Over 3,000 applications have been reported in the open literature, and many commercial projects employ the SOM as the tool for tackling real-world problems.This volume is a complete record of the Third Workshop on Self-Organising Maps, which brought together the leading international researchers and users for an intensive three-day meeting in Lincoln, UK, and so represents the very latest developments in both theory and application of SOMs and associated approaches.Topics covered include:- Pattern recognition and data clustering;- Unsupervised learning using SOMs and other techniques;- Data visualisation;- Commercial and financial applications;- Multi-dimensional signal processing;- Component and system design and condition monitoring.

This book constitutes the thoroughly refereed post-proceedings of the Third International Workshop on Graphics Recognition, GREC'99, held in Jaipur, India in September 1999.The 31 revised full papers presented were carefully reviewed and selected for inclusion in the volume. The book is divided in topical sections on vectorization, maps and geographic documents, graphic document analysis, graphic symbol and shape recognition, engineering drawings and schematics, and performance evaluation.

Presently, in our world, visual information dominates. The turn of the millenium marks the age of visual information systems. Enabled by picture sensors of all kinds turning digital, visual information will not only enhance the value of existing information, it will also open up a new horizon of previously untapped information sources. There is a huge demand for visual information access from the consumer. As well, the handling of visual information is boosted by the rapid increase of hardware and Internet capabilities. Advanced technology for visual information systems is more urgently needed than ever before: not only new computational methods to retrieve, index, compress and uncover pictorial information, but also new metaphors to organize user interfaces. Also, new ideas and algorithms are needed which allow access to very large databases of digital pictures and videos. Finally we should not forget new systems with visual interfaces integrating the above components into new types of image, video or multimedia databases and hyperdocuments. All of these technologies will enable the construction of systems that are radically different from conventional information systems. Many novel issues will need to be addressed: query formulation for pictorial information, consistency management thereof, indexing and assessing the quality of these systems. Historically, the expression Visual Information Systems can be understood either as a system for image information or as visual system for any kind information.

Multimodal Interfaces represents an emerging interdisciplinary research direction and has become one of the frontiers in Computer Science. Multimodal interfaces aim at efficient, convenient and natural interaction and communication between computers (in their broadest sense) and human users. They will ultimately enable users to interact with computers using their everyday skills. These proceedings include the papers accepted for presentation at the Third International Conference on Multimodal Interfaces (ICMI 2000) held in Beijing, China on 1416 O ctober 2000. The papers were selected from 172 contributions submitted worldwide. Each paper was allocated for review to three members of the Program Committee, which consisted of more than 40 leading researchers in the field. Final decisions of 38 oral papers and 48 poster papers were made based on the reviewers' comments and the desire for a balance of topics. The decision to have a single track conference led to a competitive selection process and it is very likely that some good submissions are not included in this volume. The papers collected here cover a wide range of topics such as affective and perceptual computing, interfaces for wearable and mobile computing, gestures and sign languages, face and facial expression analysis, multilingual interfaces, virtual and augmented reality, speech and handwriting, multimodal integration and application systems. They represent some of the latest progress in multimodal interfaces research.

The Fifth International Colloquium on Grammatical Inference (ICGI-2000) was heldinLisbononSeptember11-13th,2000.ICGI-2000wasthe?fthinaseriesof successfulbiennialinternationalconferencesintheareaofgrammaticalinference. Previous conferences were held in Essex, U.K.; Alicante, Spain; Montpellier, France; and Ames, Iowa, USA. This series of meetings seeks to provide a forum for the presentation and discussion of original research on all aspects of grammatical inference. Gram- tical inference, the process of inferring grammar from given data, is a ?eld that is not only challenging from a purely scienti?c standpoint but also ?nds many applications in real world problems. Despitethefactthatgrammaticalinferenceaddressesproblemsinarelatively narrow area, it uses techniques from many domains, and intersects a number of di?erent disciplines. Researchers in grammatical inference come from ?elds as diverse as machine learning, theoretical computer science, computational ling- stics, pattern recognition and arti?cial neural networks. From a practical standpoint, applications in areas such as natural language acquisition, computational biology, structural pattern recognition, information retrieval, text processing and adaptive intelligent agents have either been - monstrated or proposed in the literature. ICGI-2000 was held jointly with CoNLL-2000, the Computational Natural Language Learning Workshop and LLL-2000, the Second Learning Language in LogicWorkshop.Thetechnicalprogramincludedthepresentationof24accepted papers (out of 35 submitted) as well as joint sessions with CoNLL and LLL. A tutorial program organized by Gabriel Pereira Lopes took place after the meetings and included tutorials by Raymond Mooney, Gregory Grefenstette, Walter Daelemans, Ant onio Ribeiro, Joaquim Ferreira da Silva, Gael Dias, Nuno Marques, VitorRossio, Jo aoBalsaandAlexandreAgostini.Thejointrealization of these events represents a unique opportunity for researchers in these related ?elds to interact and exchange ideas."

Biometric identity verification (BIV) offers a radical alternative to passports, PIN numbers, ID cards and driving licences. It uses physiological or behavioural characteristics such as fingerprints, hand geometry, and retinas to check a person's identity. It is therefore much less open to fraudulent use, which makes it ideal for use in voting systems, financial transactions, benefit payment administration, border control, and prison access.This is the first book to provide business readers with an easy-to-read, non-technical introduction to BIV systems. It explains the background and then tells the reader how to get their system up and running quickly. It will be an invaluable read for practitioners, managers and IT personnel - in fact for anyone considering, or involved in, implementing a BIV system.Julian Ashbourn was one of the pioneers in integrating biometric technology and has provided input into many prototype BIV systems around the world.

The papers contained in this volume were presented at the 12th Annual S- posium on Combinatorial Pattern Matching, held July 1{4, 2001 at the Dan Panorama Hotel in Jerusalem, Israel. They were selected from 35 abstracts s- mitted inresponse to the call for papers. In addition,there were invited lectures by Aviezri Fraenkel (Weizmann Institute of Science), Zvi Galil (Columbia), Rao Kosaraju (Johns Hopkins University), and Uzi Vishkin(Technion and U. Ma- land). This year the call for papers invited short (poster) presentations. They also appear in the proceedings. Combinatorial Pattern Matching (CPM) addresses issues of searching and matching strings and more complicated patterns such as trees, regular expr- sions,graphs,pointsets, andarrays,invariousformats.Thegoalistoderiven- trivialcombinatorialproperties ofsuchstructures andtoexploitthese properties in order to achieve superior performance for the corresponding computational problems. On the other hand, an important aim is to analyze and pinpoint the properties and conditions under which searches can not be performed e ciently. Overthepastdecadeasteady ?owofhighqualityresearch onthissubject has changed a sparse set of isolated results into a full-? edged area of algorithmics. This area is continuing to grow even further due to the increasing demand for speed and e ciency that stems from important applications such as the World Wide Web, computational biology, computer vision, and multimedia systems. These involverequirements forinformationretrieval inheterogeneous databases, data compression, and pattern recognition. The objective of the annual CPM gathering is to provide an international forum for the presentation of research results in combinatorial pattern matching and related applications.

This book constitutes the refereed proceedings of the 4th International Workshop on Visual Form, IWVF-4, held in Capri, Italy, in May 2001.
The 66 revised full papers presented together with seven invited papers were carefully reviewed and selected from 117 submissions. The book covers theoretical and applicative aspects of visual form processing. The papers are organized in topical sections on representation, analysis, recognition, modelling and retrieval, and applications.

The papers contained in this volumewere presented at the 11thAnnual Sym- sium on CombinatorialPattern Matching, held June 21-23, 2000 at the Univ- sit edeMontr eal. They were selected from 44 abstracts submitted in response to the call for papers. In addition, there were invited lectures by Andrei Broder (AltaVista), Fernando Pereira (AT&T Research Labs), and Ian H. Witten (U- versity of Waikato). The symposium was preceded by a two-day summer school set up to - tract and train young researchers. The lecturers at the school were Greg Butler, ClementLam, andGusGrahne: BLAST Howdoyousearchsequencedatabases?, DavidBryant: Phylogeny, Ra aeleGiancarlo: Algorithmicaspectsof speech rec- nition, Nadia El-Mabrouk: Genome rearrangement, LaxmiParida: Flexib- pattern discovery, and Ian H. Witten: Adaptive text mining: inferring structure from sequences. Combinatorial Pattern Matching (CPM) addresses issues of searching and matching strings and more complicated patterns such as trees, regular expr- sions graphs, point sets, and arrays. The goal is to derive non-trivial combi- torial properties of such structures and to exploit these properties in order to achieve superior performance for the corresponding computational problems. Over recent years a steady ?ow of high-quality research on this subject has changed a sparse set of isolated results into a fully-?edged area of algorithmics

The need to improve communication between humans and computers has been instrumental in de ning new modalities of communication, and new ways of interacting with machines. Gestures can convey information for which other modalities are not e cient or suitable. In natural and user-friendly interaction, gesturescanbeused, asasinglemodality, orcombinedinmultimodalinteraction schemes which involvespeech, or textual media. Speci cation methodologiescan be developed to design advanced interaction processes in order to de ne what kind of gestures are used, which meaning they convey, and what the paradigms of interaction are. Research centred on gesture interaction has recently provided signi cant technologicalimprovements, in particular: gesture capture and tra- ing (from video streams or other input devices), motion recognition, motion generation, and animation. In addition, active research in the elds of signal processing, pattern recognition, arti cial intelligence, and linguistics is relevant to the areas covered by the multidisciplinary research on gesture as a means of communication. Resulting fromathree-dayinternationalworkshopin Gif-sur-Yvette, France, with 80 participants from ten countries all over the world, this book presents contributions on gesture under the focus of human-computer communication. The workshop was run by Universit e Paris Sud, Orsay, on the lines of GW'96 at York University, UK, and GW'97 at Bielefeld University, Germany. Its purpose was to bring together scientists from researchand industrial organisationswo- ing on all aspects of gesture modelling and interaction. The book is organised in sixsections, coveringhumanperceptionandproductionofgesture, gestureloc- isation and movement segmentation, vision-based recognition and sign language recognition, gesture synthesis and animation, and multimodality.

The AMDO 2000 workshop took place at the Universitat de les Illes Balears (UIB) on 7-9 September 2000, sponsored by the International Association for Pattern Recognition Technical Committee, the European Commission by - man Potential Program:High Level Scienti?c Conferences and the Mathematics andComputer Science DepartmentofUIB. Thesubject ofthe workshopwas- goingresearchinarticulatedmotiononthesequenceofimagesandsophisticated models for deformable objects. The goals of these areas are to understand and interpret object motion around complex objects that we can ?nd in sequences of images in the real world. These topics (geometry and physics of deformable models, motion analysis, articulated models and animation, visualization of - formable models, 3D recovery from motion, single or multiple human motion analysis and synthesis, applications of deformable models and motion analysis, etc. ) are interesting examples of how research can be used to solve more general problems. Another objective of this workshop was to relate ?elds using c- puter graphics, computer animation or applications in several disciplines c- bining synthetic and analytical images. In this regard it is of particular interest to encouragelinksbetweenresearchersinareasofcomputervisionandcomputer graphics who have common problems and frequently use similar techniques. The workshop included four sessions of presented papers and two tutorials. Invited speakers treating various aspects of the topics were: Y. Aloimonos from the Computer Vision Laboratory, Center for Automation Research, University of Maryland, USA, G. Medioni from the Institute for Robotics and Intelligent S- tems,UniversityofSouthernCalifornia,USA,andR. Boulic,Adjointscienti?que from the Swiss Federal Institute of Technology Lausanne, Switzerland. September 2000 H. -H. Nagel and F. J.

This book constitutes the thoroughly refereed post-workshop proceedings of the International Workshop on Vision Algorithms held in Corfu, Greece in September 1999 in conjunction with ICCV'99.
The 15 revised full papers presented were carefully reviewed and selected from 65 submissions; each paper is complemented by a brief transcription of the discussion that followed its presentation. Also included are two invited contributions and two expert reviews as well as a panel discussion. The volume spans the whole range of algorithms for geometric vision. The authors and volume editors succeeded in providing added value beyond a mere collection of papers and made the volume a state-of-the-art survey of their field.

This book constitutes the joint refereed proceedings of the 8th International Workshop on Structural and Syntactic Pattern Recognition and the 3rd International Workshop on Statistical Techniques in Pattern Recognition, SSPR 2000 and SPR 2000, held in Alicante, Spain in August/September 2000. The 52 revised full papers presented together with five invited papers and 35 posters were carefully reviewed and selected from a total of 130 submissions. The book offers topical sections on hybrid and combined methods, document image analysis, grammar and language methods, structural matching, graph-based methods, shape analysis, clustering and density estimation, object recognition, general methodology, and feature extraction and selection.

For many years, Arti?cial Intelligence technology has served in a great variety of successful applications. AI researchand researchershave contributed much to the vision of the so-called Information Society. As early as the 1980s, some of us imagined distributed knowledge bases containing the explicable knowledge of a company or any other organization. Today, such systems are becoming reality. In the process, other technologies have had to be developed and AI-technology has blended with them, and companies are now sensitive to this topic. TheInternetandWWWhaveprovidedtheglobalinfrastructure,whileatthe same time companies have become global in nearly every aspect of enterprise. This process has just started, a little experience has been gained, and therefore it is tempting to re?ect and try to forecast, what the next steps may be. This has given us one of the two main topics of the 23rd Annual German Conference on Arti?cial Intelligence (KI-99)held at the University of Bonn: The Knowledge Society. Two of our invited speakers, Helmut Willke, Bielefeld, and Hans-Peter Kriegel, Munich, dwell on di?erent aspects with di?erent perspectives. Helmut Willke deals with the concept of virtual organizations, while Hans-Peter Kriegel applies data mining concepts to pattern recognitiontasks.The three application forums are also part of the Knowledge Society topic: "IT-based innovation for environment and development", "Knowledge management in enterprises", and "Knowledgemanagementinvillageandcityplanningoftheinformationsociety".

Many theoretical and experimental studies have shown that a multiple classi?er system is an e?ective technique for reducing prediction errors [9,10,11,20,19]. These studies identify mainly three elements that characterize a set of cl- si?ers: -Therepresentationoftheinput(whateachindividualclassi?erreceivesby wayofinput). -Thearchitectureoftheindividualclassi?ers(algorithmsandparametri- tion). - The way to cause these classi?ers to take a decision together. Itcanbeassumedthatacombinationmethodise?cientifeachindividualcl- si?ermakeserrors'inadi?erentway',sothatitcanbeexpectedthatmostofthe classi?ers can correct the mistakes that an individual one does [1,19]. The term 'weak classi?ers' refers to classi?ers whose capacity has been reduced in some way so as to increase their prediction diversity. Either their internal architecture issimple(e.g.,theyusemono-layerperceptronsinsteadofmoresophisticated neural networks), or they are prevented from using all the information available. Sinceeachclassi?erseesdi?erentsectionsofthelearningset,theerrorcorre- tion among them is reduced. It has been shown that the majority vote is the beststrategyiftheerrorsamongtheclassi?ersarenotcorrelated.Moreover, in real applications, the majority vote also appears to be as e? cient as more sophisticated decision rules [2,13]. Onemethodofgeneratingadiversesetofclassi?ersistoupsetsomeaspect ofthetraininginputofwhichtheclassi?erisrather unstable. In the present paper,westudytwodistinctwaystocreatesuchweakenedclassi?ers;i.e.learning set resampling (using the 'Bagging' approach [5]), and random feature subset selection (using 'MFS', a Multiple Feature Subsets approach [3]). Other recent and similar techniques are not discussed here but are also based on modi?cations to the training and/or the feature set [7,8,12,21].

This volume constitutes the refereed proceedings of the 8th International Conference on Computer Analysis of Images and Patterns, CAIP'99, held in Ljubliana, Slovenia in September 1999. The 47 revised full papers presented were selected from 120 submissions. Also included are two invited contributions and 27 posters. The papers are organized in sections on color; image processing; image databases; image compression and watermarking; object and face recognition; classification and fitting; 3D-reconstruction and shape representation; motion, range image registration; applications; shape from shading, texture and stereo; real-time tracking; panoramic images; grouping; image rendering; and alignment and matching.

Ten years ago, the inaugural European Conference on Computer Vision was held in Antibes, France. Since then, ECCV has been held biennially under the auspices of the European Vision Society at venues around Europe. This year, the privilege of organizing ECCV 2000 falls to Ireland and it is a signal honour for us to host what has become one of the most important events in the calendar of the computer vision community. ECCV is a single-track conference comprising the highest quality, previously unpublished, contributed papers on new and original research in computer vision. This year, 266 papers were submitted and, following a rigorous double-blind review process, with each paper being reviewed by three referees, 116 papers were selected by the Programme Committee for presentation at the conference. The venue for ECCV 2000 is the University of Dublin, Trinity College. - unded in 1592, it is Ireland's oldest university and has a proud tradition of scholarship in the Arts, Humanities, and Sciences, alike. The Trinity campus, set in the heart of Dublin, is an oasis of tranquility and its beautiful squares, elegant buildings, and tree-lined playing- elds provide the perfect setting for any conference.

Recently, there has been an increased interest in the research and development of techniques for components of complete document analysis systems. In recognition of this trend, a series of workshops on Document Analysis Systems commenced in 1994, under the leadership of Henry Baird. The first workshop, held in Kaiserslautern, Germany, in October, 1994, was chaired by Andreas Dengel and Larry Spitz. The second workshop on Document Analysis Systems was held in Malvern, PA, USA, in October, 1996, chaired by Jonathan J. Hull and Suzanne Liebowitz Taylor. The DAS workshop has been one of the most prestigious technical meetings, bringing together a large number of scientists and engineers from all over the world to express their innovative ideas and report on their latest achievements in the area of document analysis systems. The papers in this special book edition were rigorously selected from the Third IAPR Workshop on Document Analysis Systems (DAS'98), held in Nagano, Japan, on 4 - 6 November 1998. It is worth mentioning that the papers were chosen for their original and substantial contributions to the workshop theme and this special book edition. From among the 53 papers that were presented by authors from 11 countries at the DAS'98 after critical reviews by at least three experts, we carefully selected 29 papers for this special book edition. Most of the contributions in this edition have been expanded or extensively revised to include helpful discussions, suggestions, or comments made during the workshop.

This book constitutes the thoroughly refereed post-proceedings of the International Workshop on Graph Transformation with Industrial Relevance, AGTIVE'99, held in Kerkrade, The Netherlands, in June 1999.
The 28 revised full papers presented went through an iterated process of reviewing and revision. Also included are three invited papers, 10 tool demonstrations, a summary of a panel discussion, and lists of graph transformation systems and books on graph transformations. The papers are organized in sections on modularization concepts, distributed systems modeling, software architecture: evolution and reengineering, visual graph transformation languages, visual language modeling and tool development, knowledge modeling, image recognition and constraint solving, process modeling and view integration, and visualization and animation tools.

The papers contained in this volume were presented at the Tenth Annual S- posium on Combinatorial Pattern Matching, held July 22 { 24, 1999 at the University of Warwick, England. They were selected from 26 abstracts subm- ted in response to the call for papers. In addition, invited lectures were given by JoanFeigenbaumfromAT&TLabsResearch(Massivegraphs: algorithms, app- cations, and open problems) and David Jones from the Department of Biology, University of Warwick (Optimizing biological sequences and protein structures using simulated annealing and genetic algorithms). The symposium was preceded by a two-day summer school set up to attract and train young researchers. The lecturers of the school were Alberto Ap- tolico (Computational Theories of Surprise), Joan Feigenbaum (Algorithmics of network-generatedmassivedatasets), Leszek Gasieniecand PaulGoldberg (The complexity of gene placement), David Jones (An introduction to computational molecularbiology), Arthur Lesk (Structuralalignmentandmaximalsubstructure extraction), Cenk Sahinalp(Questformeasuringdistancebetweenstrings: exact, approximate, and probabilistic algorithms), and Jim Storer. Combinatorial Pattern Matching (CPM) addresses issues of searching and matching strings and more complicated patterns such as trees, regular expr- sions, graphs, point sets, and arrays. The goal is to derive non-trivial combi- torial properties of such structures and to exploit these properties in order to achieve superior performance for the corresponding computational problems. Over recent years, a steady ?ow of high-quality research on this subject has changed a sparse set of isolated results into a fully-?edged area of algorithmics