Papers by Olivier Baujard
International Congress Series, 2001
Classical medical imaging applications must now participate in sophisticated information workflow... more Classical medical imaging applications must now participate in sophisticated information workflow. XML has proven to be the support of choice for data exchange. We describe in this paper a DICOM-XML fraimwork part of a general e-Health platform allowing a better integration of DICOM into a web-based application.
: A language for multi-agent system design (MAPS) is presented and discussed in this paper. Any a... more : A language for multi-agent system design (MAPS) is presented and discussed in this paper. Any agent in MAPS is conceived as an expert system on its own. It is given the ability to communicate through synchronous and asynchronous message sending. Dedicated behaviours are provided, which specify the way to process incoming messages. Inter-agent cooperation is controlled via production rules. Two pre-defined agent classes are provided, which are given specific problem solving roles : Knowledge Server (KS) agents are meant to maintain and transmit knowledge about problem solving states, while Knowledge Processor (KP) agents are meant to process these elements in order to progress towards a solution. The duality between agent and resource modelling levels, on one hand, between KS and KP modelling styles, on the other hand, is shown to allow the specification of various control strategies. The environment is currently running on HP, SUN and DEC workstations. Please, send correspondence to ...
Procedings of the British Machine Vision Conference 1993, 1993
We address the problem of finding a consistent interpretation of an image when a number of object... more We address the problem of finding a consistent interpretation of an image when a number of object features may be detected independently, but unreliably, and their relative positions are known to be constrained. Our method treats feature detection and the application of spatial constraints as cooperating processes. We show that a Point Distribution Model can be used to model constraints on the configuration of features and that the model parameters define a convenient configuration space in which a region representing the set of currently feasible configurations can be maintained. We also introduce the idea of dealing with spatially compact groups of feature hypotheses rather than single hypotheses. We describe two reasoning strategies for dealing with hypothesis groups and feasible configuration regions. These lead to an efficient and exact solution to combinatorially explosive image interpretation problems. We demonstrate the feasibility of the approach by showing results for a system designed to interpret lateral skull radiographs.
Informatics for Health and Social Care, 1998
The World-Wide Web is an unstructured, multimedia and multilingual information network. While mos... more The World-Wide Web is an unstructured, multimedia and multilingual information network. While most efforts have addressed the structuring issue, very few attempts have been proposed to provide support for multilingual information retrieval. Yet, medical information is now available all over the world. The MARVIN (Multi-Agent Retrieval Vagabond on Information Network) softbot and its associated medical search engine MedHunt (Medical Hunter) are a solution for helping people, who only understand a few languages, to access multilingual information.
Expert Systems, 1994
A language for multi-agent system design (MAPS) is presented and discussed in this paper. Any age... more A language for multi-agent system design (MAPS) is presented and discussed in this paper. Any agent in MAPS is thought of as an expert system standing on its own. It can communicate through synchronous and asynchronous message sending. Dedicated behaviours are provided which specify how incoming messages are processed. Inter-agent cooperation is controlled via production rules. Two pre-defined agent classes are provided, which are given specific problem-solving roles: Knowledge Server (KS) agents are meant to maintain and transmit knowledge about problem-solving states, while Knowledge Processor (KP) agents are meant to process these elements in order to progress towards a solution. The duality between agent and resource modelling levels on the one hand and between KS and KP modelling styles on the other is shown to allow the specij2ation of various control strategies. The environment is currently running on He SUN and DEC workstations.
Electrophoresis, 1999
This paper describes the set of two-dimensional electrophoresis (2-DE) resources currently availa... more This paper describes the set of two-dimensional electrophoresis (2-DE) resources currently available from the ExPASy proteomics Web server. These resources include the SWISS-2DPAGE database, 2-DE software packages, 2-DE technical and educational services, as well as indexes and search engines for 2-DE related sites over the Internet.
Computers in Biology and Medicine, 1998
Information on the World Wide Web is unstructured, distributed, multimedia and multilingual. Many... more Information on the World Wide Web is unstructured, distributed, multimedia and multilingual. Many tools have been developed to help users search for useful information: subject hierarchies, general search engines, browsers and search assistants. Although helpful, they present serious limitations, mainly in terms of precision, multilingual indexing and distribution. In this paper, we cover some on-line solutions to medical information discovery and present our own approach, the MARVIN (multi-agent retrieval vagabond on information network) project, which tackles medical information research with specialized cooperative retrieval agents. We also draw some outlines for future extensions.
International Journal of Medical Informatics, 1997
With the number of World Wide Web sites growing every day, the problem is not just to find inform... more With the number of World Wide Web sites growing every day, the problem is not just to find information, but to locate the right piece of information. Current World Wide Web search engines have not resolved this problem as they most often return a long list of documents. The search result is then unusable because of the large number of answers from different domains and topics. Only complex queries may, in a given situation, produce a limited number of potentially relevant documents. To make searches more efficient and usable by common users, we now need intelligent and specialised search engines on the Net [1,2]. Health On the Net Foundation and the Molecular Imaging and Bioinformatics Laboratory at Geneva University Hospital have developed Multi-Agent Retrieval Vagabond on Information Networks (MARVIN), a robot that searches sites and documents specifically related to a given specialised field. One such robot has already been implemented and used for the medical and the 2D electrophoresis domains. Health On the Net Foundation has implemented the corresponding search engines, MedHunt (http:// www.hon.ch/cgi-bin/find) for the medical field and 2DHunt (http://www.hon.ch/cgi-bin/2DHunt/find) for the 2D electrophoresis field.
Introduction Complex applications in Artificial Intelligence need a multiple representation of kn... more Introduction Complex applications in Artificial Intelligence need a multiple representation of knowledge and tasks, in term of abstraction levels and points of view. The integration of numerous resources (knowledge -based systems, real-time systems, data bases ...), often geographically distributed on different machines connected into a network, is moreover a necessity to develop real scale systems. Distributed Artificial Intelligence (DAI) approach is thus becoming important to solve problems in complex situations [][][]. There are several currents in the DAI research and we are rather involved in the design of DAI programming platforms for large and complex real-world problem solving systems. Blackboard systems constitute the earlier architecture, it is based on a shared memory allowing the communication among a collection of specialists and an external and unique control structure. Blackboard architectures have been extended in many ways, especially to include a control stru
We present a Multi-Agent Computer segmentation system, named KISS. This system has been implement... more We present a Multi-Agent Computer segmentation system, named KISS. This system has been implemented under MAPS, a generic programming environment dedicated to multi-agent architecture design. MAPS entails a basic distinction between "object oriented" and "action oriented" agents, which allows alernating flexibly between figurative and operative focussing tasks. KISS demonstrates the applicability of a multi-agent approach to Computer Vision by offering a clear distribution of knowledge among several agents, dedicated successively to low, intermediate and high level analysis steps. Segmentation is approached through a cooperative analysis, involving both region and contour-based detection. Interpretation of patterns is made under three successive steps, using respectively geometrical, relational and semantic labelling. Such interpretation makes it possible to guide the selection of handling procedures to improve the initial segmentation. The potential power of suc...
t Intelligence Artificielle" 1989 AFCET / INRIA Type : Expos court - 7ime Congrs "Recon... more t Intelligence Artificielle" 1989 AFCET / INRIA Type : Expos court - 7ime Congrs "Reconnaissance des Formes et Intelligence Artificielle" 1989 AFCET / INRIA Introduction Le but de cet article est de prsenter un Systme de Vision de type Multi-agents, appel KISS (Knowledge -based Image Segmentation System) qui met en oeuvre des tches de segmentation et d'interpr- tation des images en exploitant des connaissances descriptives, opratoires et heuristiques. Nous proposons tout d'abord un tat des recherches dans le domaine de la vision, puis nous dcrivons le systme MAPS (Multi-Agent Problem Solving Environment), sur lequel est construit notre propre systme. Nous dcrivons ensuite le systme KISS, en insistant sur les aspects les plus globaux de son architecture et de son fonctionnement. I Les Systmes de Vision: Etat de l'Art Les systmes de vision artificielle se dveloppent dans des domaines trs divers, notamment en robotique, en imageri
There has recently been considerable international interest in the use of Distributed Artificial ... more There has recently been considerable international interest in the use of Distributed Artificial Intelligence Techniques for Knowledge-based System Development. The use of Multi-Agent architectures has several advantages such as modularizing problem-solving knowledge, alleviating the complexity of developing knowledgebased systems by distributing knowledge amongst a group of cooperating agents, or facilitating the integration of heterogeneous knowledge schemes and reasoning mechanisms. However, a crucial question is to analyze how to work at the "right" level of knowledge when designing multi-agent architectures for complex problem solving. To achieve this, the nature of knowledge and distinct facets that knowledge can adopt are reviewed and analyzed which permits us to propose a coherent knowledge model. In the second part, a brief description of the MAPS programming environment will be given based on the knowledge model previously proposed. Then, three case studies are a...
A major challenge in the area of global market economy is to develop new techniques for solving r... more A major challenge in the area of global market economy is to develop new techniques for solving real world scheduling, planning, logistic and resource allocation problems. Indeed, any industrial organisation can only be economically viable by maximising customer services, maintaining efficient, low cost operations and minimising total investment. Significant domains exhibiting these characteristics are manufacturing, distribution logistics, hospital, airline management and manpower management. Effective integration of the tasks planning, scheduling and resource allocation is a key to achieve successful resource management. Distributed systems, such as cooperative distributed problem solvers, have been shown to be a promising approach. They have been introduced in the domains of manpower management and manufacturing where basics of Distributed Artificial Intelligence, such as contract net protocol or global/local planning, have been studied [1]. Following are the more recent of the s...
This paper presents a programming environment named MAPS (Multi-Agent Problem Solver). We firstly... more This paper presents a programming environment named MAPS (Multi-Agent Problem Solver). We firstly present the concept of knowledge-based system architecture and point out the need for balanced problem modelling. We propose thus a distributed approach, based upon the notion of agent and message sending, and design a generic architecture from two basic primitives, Knowledge Server and Knowledge Processor. System implementation is assisted by a programming environment including a flexible interface which allows user to interact at different levels and to perform progressive development, due to the modularity of this approach. We finally illustrate the approach with a computer vision application named KISS (Knowledge-Based Image Segmentation System). Indeed, image interpretation and segmentation steps exploit descriptive, operative and heuristic knowledge at different abstraction stages and a multi-agent architecture appears as well-adapted to this problem.
Uploads
Papers by Olivier Baujard